Historians Fallacies Toward A Logic Of Historical Thought

Historians*
Fallacies

Toward a Logic of
Historical Thought

HISTORIANS'
FALLACIES

Toward a Logic of Historical Thought

HISTORIANS'
FALLACIES

Toward a Logic of Historical Thought
by David Hackett Fischer

HARPER PERENNIAL
NEW YORK • LONDON • TORONTO • SYDNEY • NEW DELHI • AUCKLAND

HARPER PERENNIAL

The quotation on page 20 is from "Burnt Norton" in Four Quartets by T. S. Eliot,

reprinted by permission of Harcourt, Brace It World, Inc.
The lines by Robert Frost on page 130 are from "The Lesson for Today" in Complete

Poems by Robert Frost, reprinted by permission of Holt, Rinehart and Winston, Inc.

hvtduans' fallacies. Copyright © 1970 by David Hackett Fischer. All rights reserved.
Printed in the United Sulci of America. No part of this book may be used or reproduced
in any manner whatsoever without written pcrmiasoo except in the case of brief quota-
tions embodied in critical articles and reviews. For information address Harper k Row,
Publishers, Inc.. 10 East 53d Street. New York, N.Y. 10022

First harper torch book edition published 1970
ISBN: 0-06-131 545-1

11 12 13 RRD(H) 50 49 48 47 46

For Margaretta Frederick

CONTENTS

Preface ix

Introduction xv

PART I INQUIRY

Chapter I Fallacies of Question-Framing 3

Chapter II Fallacies of Factual Verification 40

Chapter III Fallacies of Factual Significance 64

PART II EXPLANATION

Chapter IV Fallacies of Generalization 103

Chapter V Fallacies of Narration 131

Chapter VI Fallacies of Causation 164

Chapter VII Fallacies of Motivation 187

Chapter VIII Fallacies of Composition 216

Chapter IX Fallacies of False Analogy 243

PART III ARGUMENT

Chapter X Fallacies of Semantical Distortion 263

Chapter XI Fallacies of Substantive Distraction 282

Conclusion 307

Index 319

Index of Fallacies 337

PREFACE

The poor condition of the logical analysis of history is
shown by the fact that neither historians, nor methodolo-
gists of history, but rather representatives of very unrelated
disciplines have conducted the authoritative investigations
into this important question.

—Max Weber

The problem of locating a logic of historical thought cuts across several
separate academic disciplines. For that reason, perhaps, it has been
satisfactorily studied by none of them. Many professional logicians will
refuse to recognize it as a logical problem at all. They have been at some
pains to show that their subject is an intellectual discipline in its own
right — even the intellectual discipline. They are customarily committed
to a search for the logic of thought about everything in general, and
nothing in particular. One philosopher, Stephen Toulmin, has proposed
a different strategy: the refinement of a plurality of field-related logics
which are designed to promote special kinds of inquiry. "Not only will
logic have to become more empirical," Toulmin writes, "it will inevitably
tend to become more historical. To think up new and better methods of
arguing in any field is to make a major advance, not just in logic, but
in the substantive field itself." 1 Toulmin's manifesto was issued more than
ten years ago. Few logicians have responded to his call. Most of them
are still moving in the opposite direction.

If there is a field-related logic of historical thought, then working
historians must help to find it. But they have contributed little of con-
sequence in the past forty years. Their articles and books on the nature
of history tend to degenerate into mere exhortations, or manuals on the
mechanics of citation, or metahistorical mumbo-jumbo. Many academic
historians regard methodological and logical problems with suspicion
and even hostility. Incredibly, the word "logic" is often a pejorative,

1. Stephen Toulmin, The Uses of Argument (Cambridge, 1958), p. 257.

X

PREFACE

in their professional parlance. Sometimes it is used as a synonym for
determinism by embattled antideterminists. A Scottish historian writes,
"No doubt it was 'in the logic of history' that England should endeavor
to absorb Scotland. None the less, the first English attempt at a whole-
sale conquest resulted from two dynastic accidents. . . . " 2

A different usage carries an equally disagreeable connotation.
"Logic" is sometimes understood by historians as a treacherous tangle
of disjunctive snares and syllogistic delusions which are the last resort
of a scholar who has run short of sources. An American historian writes
contemptuously of a colleague, "He had little evidence to go upon as
yet, and so he resorted frankly to logic." Another historian describes
logic as "Machiavellian" in a recent monograph. 3

This unfortunate attitude has deep roots in historians' thought. It
reaches back to Carlyle's contempt for the "dead logic formula," and to
Guizot's epigram that "Nothing falsifies history more than logic." In the
past generation, the progress of this prejudice among historians has been
promoted by the influence of relativism. That great blight upon histor-
ical scholarship is remembered as a repudiation of the empirical aspira-
tions of "scientific history." It was also a revolt against reason. Carl
Becker wrote:

We have long since learned not to bother overmuch with reason and logic.
Logic was formerly visualized as something outside us, something existing
independently which, if we were willing, could take us by the hand and lead
us into the paths of truth. We now suspect that it was something the mind
has created to conceal its timidity and keep up its courage, a hocus-pocus
designed to give formal validity to conclusions we are willing to accept if
everybody else in our set will loo. If all men are mortal (an assumption),
and if Socrates was a man (in the sense assumed), no doubt Socrates must
have been mortal; but we suspect that we somehow knew all this before
it was submitted to the test of a syllogism. Logics have a way of multiplying
in response to the changes in point of view. . . . The secure foundations of
deductive and inductive logic have been battered to pieces by the ascertain-
able facts, so that we really have no choice; we must cling to the ascertain-
able facts though they slay us. 4

This is indeed an irony, in a scholar who was so radically skeptical
about "ascertainable facts." But his attitude retains its popularity, par-
ticularly among Anglo-American historians whose professional prej-
udices are powerfully reinforced by a cultural predisposition. Stanley
Baldwin boasted in the year 1937, that "one of the reasons why our

2. John Duncan Mackie. A History of Scotland (Baltimore, 1966), p. 71.

3. Wesley Frank Craven, The Legend of the Founding Fathers (Ithaca, 1965), p. 120;
Carl E. Prince, New Jersey's Jeffersonian Republicans (Chapel Hill, N.C., 1967), p. 119.

4. Carl Becker, The Heavenly City of the Eighteenth Century Philosophers (New Haven,
1932), pp. 25-26.

PREFACE

xi

people are alive and flourishing, and have avoided many of the troubles
that have fallen to less happy nations, is that we have never been guided
by logic in anything we did." 5

There is an appropriately ugly name for this prejudice: misology,
or logic-hating. Its continuing existence among academic historians
explains their failure to refine a logic of historical thought. When a
distinguished American historian openly asserts that "a good bed book
is more to be desired than another Critique of Pure Reason,'" 1 it is not
astonishing that a Critique of Historical Reason has failed to appear.

A good deal of relevant and important work has recently been done,
not by logicians or historians, but by epistemologists. Today, a special
subdiscipline of epistemology called the analytical philosophy of history
is in a flourishing condition. Any historian who wishes to understand
the nature of his own work has much to learn from it, and particularly
from two excellent new books by Arthur Danto and Morton White. 7
But the work of Danto and White and their colleagues is not very useful
in any attempt at the refinement of the thinking of working historians,
for three reasons. First, analytical philosophers of history are simply not
much interested in low problems of utility. "Today's philosopher of
history is not a metaphysical speculator, but neither is he a methodolog-
ical consultant," writes Morton White. 8 Second, historical epistemo-
logists have not been sufficiently empirical in their procedures. They
have not sufficiently attended to historical thinking as it actually happens,
or to historical problems as they actually exist. Third, epistemologists
have characteristically tried to analyze historical knowledge in terms of
something else more familiar to them. Often in terms of some epistemo-
logical abstraction which bears small resemblance to the thinking which
anybody actually does. Most analytical philosophers who have written
on the subject attempt to force historical knowledge into a formula which
the cognoscenti call the Hempelian model, or the Covering Law model,
or the Deductive Model of Explanation. This abstraction, I believe, is
seriously mistaken as an understanding of the thinking of historians, or
of social scientists, or of natural scientists for that matter. 9 But all his-

5. Quoted in L. Susan Stebbing, Thinking to Some Purpose (Baltimore, 1961), p. 17.
Winston Churchill similarly remarked that "Logic is a poor guide compared with custom.
Logic . . . has proved fatal to parliamentary government." Randolph S. Churchill, Winston
Churchill, The Young Statesman (Boston, 1967), p. 5.

6. Charles F. Mullett, in The American Historical Review 72 (1966): 186.

7. Arthur Danto, Analytical Philosophy of History (Cambridge, 1965); Morton White,
The Foundations of Historical Knowledge (New York, 1965).

8. White, The Foundations of Historical Knowledge, p. 2.

9. For an extended critique by a working social scientist, see Eugene Meehan, Explana-
tion in the Social Sciences: A System Paradigm (Homewood, 111., 1968). See also
chapter 4, below.

xii

PREFACE

torians are made to lie down in this Procrustean bed, even if their heads
must be removed to make them fit. A good many intelligent men have
wasted a great deal of time and effort to reconcile this epistemological
error with empirical facts which so obviously contradict it. The results
are often ingenious, but rarely productive. Stirrings of a new spirit are
slowly beginning to appear, but the Hempelian model, in modified forms,
retains its popularity.

There are other schools of historical epistemology, which are
equally unsatisfactory. One of them is organized around the central
idea that to write history is to tell a story. This, I think, is partly true for
some historians, but entirely false for others, and insufficient for all.
I shall argue in the following chapters that history-writing is not story-
telling but problem solving. Sometimes the solution takes the form of
a story. But often (and increasingly today) a different kind of explana-
tion-strategy is adopted. 10

Still another major school of historical epistemology attempts to
force all historical thought into an idealist model which derives from
the work of R. G. Collingwood. I believe this "idea of history" to be
fallacious, and have briefly discussed it as such in Chapter 7, below."

If the analytical philosophy of history is presently in an unsatisfac-
tory condition, professional historians have nobody to blame but them-
selves. "Historians," a philosopher has fairly complained, "show an
almost pathological disinclination to commit themselves to general
statements about their work, its aims, subject matter, and methods." 12
A few of my colleagues have even argued that a "working historian"
ought not to be "talking about such matters at all." 13 Others, when asked
to explain the nature of history, are apt to respond as Fats Waller (or
maybe Louis Armstrong) did, when asked to explain the nature of jazz.
"Man," he said, "if you don't know what it is, don't mess with it."

These attitudes have taken a heavy toll of modern historiography.
Those historians who imagine themselves to be emancipated from phi-
losophy are apt, in Keynes's phrase, to be the slaves of some defunct
philosopher. If the logic and epistemology of historical thought are to
be understood, if historical and logical and epistemological thinking are
to be refined, then historians, logicians, epistemologists, and others must
work together in a spirit of mutual cooperation. Each of these proud
disciplines has much to teach the others — and much to learn as well.

10. The most extended statement of this understanding of history as story-telling appears
in W. B. Gallie, Philosophy and the Historical Understanding (New York, 1964).

11. The most influential text is R. G. Collingwood, The Idea of History (Oxford, 1946).

12. W. B. Gallie, Philosophy and the Historical Understanding, p. 53.

13. Bernard Bailyn, in "The Problems of a Working Historian: A Comment," in Sidney
Hook, ed., Philosophy and History (New York, 1963), pp. 93-94.

PREFACE

xiii

Many people have contributed to the writing of this book. My
interest in the subject was awakened ten years ago by a philosopher,
Ronald Butler, in an undergraduate course at Princeton University on
the analytical philosophy of history. More recently I have learned much
in conversation and correspondence with many colleagues — Jerold
Auerbach, Jacob Cohen, Norman Cantor, Ray Ginger, Morton Keller,
Leonard Levy, Heinz Lubasz, Eugene Meehan, and especially Marvin
Meyers who always had time to talk out a knotty problem and construc-
tive suggestions for its solution. Ramsay MacMullen kindly answered
a question on Livy, and Douglas Stewart helped with problems of Greek
historiography. George Billias gave me a chance to try out my ideas on
some of his students, and also the benefit of his own criticism. J. H.
Hexter and Arthur Danto read and criticized the manuscript, and it is
better for their suggestions. A special debt is due to my students in His-
tory 97 at Brandeis University, and particularly to Hillel Schwartz and
Eric Uslaner who voluntarily produced written critiques of the book.

The editorial staff of Harper & Row was, as ever, a model of en-
couragement and efficiency. I am especially obliged to Hugh Van Dusen
and Cynthia Merman for their help, and to Antonia Rachiele for her
intelligent and painstaking criticism of the manuscript. Brandeis Uni-
versity provided a generous grant-in-aid. Judith, my wife, made the pro-
ject possible in every other way.

D.H.F.

Wayland, Massachusetts
September, 1969

INTRODUCTION

When we run over libraries persuaded of these principles,
what havoc must we make?

— David Hume

This book begins with three related premises: first, that there is a tacit
logic of historical thought; second, that this logic can be raised to the
level of awareness; and third, that historical thinking itself can be refined
by its intelligent and purposeful application.

The logic of historical thought is not a formal logic of deductive
inference. It is not a symmetrical structure of Aristotelian syllogisms, or
Ramean dialectics, or Boolean equations. Nor is it precisely an inductive
logic, like that of Mill or Keynes or Carnap. It consists neither in in-
ductive reasoning from the particular to the general, nor in deductive
reasoning from the general to the particular. Instead, it is a process of
adductive reasoning in the simple sense of adducing answers to specific
questions, so that a satisfactory explanatory "fit" is obtained. The answers
may be general or particular, as the questions may require. History is, in
short, a problem-solving discipline. A historian is someone (anyone)
who asks an open-ended question about past events and answers it with
selected facts which are arranged in the form of an explanatory paradigm.
These questions and answers are fitted to each other by a complex pro-
cess of mutual adjustment. The resultant explanatory paradigm may
take many different forms: a statistical generalization, or a narrative,
or a causal model, or a motivational model, or a collectivized group-
composition model, or maybe an analogy. Most commonly it consists
not in any one of these components but in a combination of them.
Always, it is articulated in the form of a reasoned argument. 1

1. In this book an event is understood as any past happening. A fact is a true descriptive
statement about past events. To explain is merely to make plain, clear, or understand-
able some problem about past events, so that resultant knowledge will be useful in
dealing with future problems. An explanatory paradigm is an interactive structure of
workable questions and the factual statements which are adduced to answer them.

xvi

INTRODUCTION

To argue that there is a tacit logic of historical thinking is to assert
that every historical project is a cluster of constituent purposes, and that
each of these purposes imposes its own logical requirements upon a
thinker who adopts them. Whether the purpose at hand is to design a
proper question, or to select a responsive set of factual answers, or to
verify their factuality, or to form them into a statistical generalization
which itself becomes a fact, or whatever — it always involves the making
of purposive and procedural assumptions that entail certain logical
consequences. Every historian must learn to live within the limits which
his own freely chosen assumptions impose upon him. These assumptions
may differ radically from one historian to the next, but always they exist,
and a historian must learn to respect them. If he does not, then he will pay
a penalty in a diminution of the degree to which his purposes are attained.
No man is free from the logic of his own rational assumptions — unless
he wishes to be free from rationality itself.

Assuming that this logic of historical thought does tacitly exist,
the next question is how to raise it to the level of consciousness. In the
opinion of some intelligent men, this task is not merely difficult but
impossible. Michael Polanyi has suggested that scientists do indeed
proceed by a logic of tacit inference — but one which is only learned
through personal experience and can never be articulated. "Any attempt
to gain complete control of thought by explicit rules," he flatly declares,
"is self -contradictory, systematically misleading, and culturally destruc-
tive." 2

Polanyi's caveat would surely be correct if the object were to gain
complete control of thought. But maybe a more humble attainment is

By adduction I do not mean what Charles Sanders Peirce appears to have intended
by abduction. Peirce distinguished three kinds of reasoning. Deduction he understood
in an ordinary way as "necessary reasoning" which "starts from a hypothesis, the
truth or falsity of which has nothing to do with the reasoning." Induction he defined
in a special sense as "the experimental testing of a theory," and abduction as "the
process of forming an explanatory hypothesis." Of the latter, he wrote, "It is the only
logical operation which introduces any new idea; for induction does nothing but deter-
mine a value, and deduction merely evolves the necessary consequences of a pure
hypothesis. Deduction proves that something must be; Induction shows that something
actually is operative; Abduction merely suggests that something may be." Collected
Papers (Cambridge, 1931), V, 142, 145, 170-172. The processes which Peirce calls
abductive and inductive are combined in what I call adduction — as in fact I believe them
to be inseparably joined in historical thinking.

2. Michael Polanyi, "The Logic of Tacit Inference," Philosophy 41 (1966): 18; and
Personal Knowledge (Chicago, 1938), passim. I am indebted to Polanyi's work for the
idea of a tacit logic and for many other things, though I disagree with him on this
point.

INTRODUCTION

xvii

possible. Perhaps one might refine (not control) some kinds of thinking
by a partial articulation of some parts of this tacit logic. It seems reason-
able to expect that a man who learns much from his own experience
can also learn a little from the experience of others.

Still, the problem of locating a logic of historical thinking defies a
direct approach. Every attempt (there have been many) to storm the
citadel by a conceptual coup de main has failed of its objective. But if
a frontal assault is impossible, maybe the problem can be outflanked and
taken from behind. A historian has written suggestively that "our present
state of knowledge is one of mitigated ignorance. In such situations,
the honest enquirer always has one consolation — his blunders may be
as instructive as his successes." 3

Such is the perversity of human perceptions that a blunder is apt
to be more visible than a success. This psychological fact suggests a
crude and eccentric method, which is adopted in this book. If there is
a tacit logic of historical inquiry, then one might hope to find a tacit
illogic as well, which reveals itself in the form of explicit historical errors.
On this assumption, I have gone looking for errors in historical scholar-
ship, and then for their common denominators, in the form of false
organizing assumptions and false procedures. These common denom-
inators are called fallacies in this book. A fallacy is not merely an error
itself but a way of falling into error. It consists in false reasoning, often
from true factual premises, so that false conclusions are generated. 4

The object in the following chapters is not to compile a definitive
catalogue of historians' fallacies, which is obviously impossible. A logi-
cian, Augustus de Morgan, wisely observed that "there is no such thing
as a classification of the ways in which men may arrive at an error: it
is much to be doubted whether there ever can be." 1 ' Surely, there can be
no conclusive and comprehensive classification. Nevertheless, a list of
common fallacies — however crude and incomplete — may serve a useful
purpose in two respects. First, it may clearly indicate a few mistaken
practices that are not sufficiently recognized as such. Second, it might

3. Alan Simpson, The Wealth of the Gentry, 1540-1660 (Chicago, 1961), p. 21.

4. This definition of fallacy conforms to the third meaning of the term in Alfred
Sidgwick, Fallacies (London, 1883). It should be clearly distinguished from several
others. The literal Latin meaning of fallax suggests a deliberate deception. This, of
course, does not apply to any of the following fallacies, all of which are self-deceptions.
A fallacy has also been defined, in Jeremy Bentham's phrase, as a "vulgar error," or a
common misconception. This is too broad for our purposes. Sometimes, fallacies are
also understood as violations of the formal rules of deductive inference. But this is
irrelevant here.

5. Augustus de Morgan, Formal Lope, 1847 (London, 1926), p. 276.

xviii

INTRODUCTION

operate as a heuristic device for the discovery of a few constructive
rules of reason.

The reader might protest that this method is like telling a traveler
how to get from Boston to New York by describing in detail the roads
which won't take him there. If this were in fact our purpose, the project
would be absurd. But it is something different. The object is not to de-
scribe the ways in which a traveler might get lost, but rather to identify
a few common ways in which others have actually gone wrong. For a
traveler from Boston to New York there are an infinity of wrong routes
and a plurality of right ones. But real travelers who actually get lost
tend to do so in a few finite ways. The Public Roads Commission does
not need to put up signs everywhere but only at the doubtful intersections.

So it is with historical travelers, who set out toward a certain des-
tination. There are many intersections along the way. Some are simple
forks in the road. A few are baffling interchanges. The traveler's diffi-
culties are compounded by the fact that well-meaning people have put
up many mistaken signs for the convenience of passers-by. The signs
say, "A, this way, seven miles," but point squarely to noX-A.

The purpose of this book is, first, to pull down some of these wrong
signs. The fact that it cannot pull down all wrong signs, or that pulling
down is a destructive act, cannot be an argument against it. Second, the
object is to put up a few crude but hopefully more correct markers at
some of the simple forks in the road. Third, it is to explore some of the
baffling interchanges in a preliminary way.

The object is emphatically not three other things. It is not to put
up signs everywhere — there isn't enough lumber and paint in the world
for that. Nor is it precisely to survey the road, which cannot be done
until we have a rough sense of its location, and which will not be done
until historiographical surveyors become a little more expert in the use
of theodolites and trigonometry. Most important, the object is not to
play traffic policeman or magistrate: it is not to flag down erring travelers
and take away their licenses. In the republic of scholarship, every citizen
has a constitutional right to get himself as thoroughly lost as he pleases.
The only purpose here is to indicate, in an advisory spirit, a few wrong
turnings which have actually been taken, and to extract from these mis-
takes a few rough rules of procedure.

Somebody once asked Thomas Edison about his rules of procedure
and received a rude reply: "Rules!" said Edison, "Hell! There ain't no
rules around here! We're tryin' to accomplish sump'n." A good many
historians, particularly of the present permissive generation, which has

INTRODUCTION

xix

made a cult of flexibility in its procedures, seem to have formed the same
idea of their own discipline. I believe that they are wrong. There are some
very strict tautological rules of historical scholarship, which are rather
like the rules of chess. When a chess player sits down to a game, he must
respect a rule which requires him to move his bishops on the diagonal.
Nobody will arrest him if he doesn't. But if he refuses to play that way,
then he isn't exactly playing chess.

There are other kinds of rules in chess, too — rough experiential
rules of thumb, such as one which urges a beginning player always to
seize the open file. He can violate this rule with impunity, if he is very
lucky, or very good. But most players, in most situations, are properly
urged to respect it. 8

I hope that a study of the tacit logic of historical thought will yield
rules of both these types. But even if not, a more precise understanding
of error itself might serve a serious and constructive scholarly purpose.
Karl Popper has suggested that science develops by a sequence of "con-
jectures and refutations." He has written that "the way in which know-
ledge progresses, and especially our scientific knowledge, is by unjustified
(and unjustifiable) anticipations, by guesses, by tentative solutions to
our problems, by conjectures. These conjectures are controlled by crit-
icism; that is, by attempted refutations, which include critical tests." 7

The fallacies in the following pages might be useful as some of these
"critical tests" to which conjectures are submitted. As the pace of intel-
lectual innovation continues to accelerate, we must develop devices
which distinguish sound innovations from unsound ones. As we become
more experimental in our thoughts and acts, we must find a way to deal
with experiments that fail. In historical scholarship, the progress of inter-
pretative revision requires a degree of critical rigor that is conspicuously
absent today.

Historians must, moreover, develop critical tests not merely for
their interpretations, but also for their methods of arriving at them.
Today, there is a good deal of hostility against method among historians,
who are apt to be contemptuous of other disciplines in which this interest
is more highly developed. Among my colleagues, it is common to believe
that any procedure is permissible, as long as its practitioner publishes
an essay from time to time, and is not convicted of a felony. The re-
sultant condition of modern historiography is that of the Jews under the
Judges: every man does that which is right in his own eyes. The fields

6. I have shamelessly stolen this simile from Abraham Kaplan, The Conduct of Inquiry
(San Francisco, 1964).

7. Karl Popper, Conjectures and Refutations (New York, 1962), p. vii.

XX

INTRODUCTION

are sown with salt, and plowed with the heifer, and there is a famine upon
the land.

It ought to be immediately apparent that some historical methods
are not as good as others, for purposes at hand. And a few methods in
common use are simply no good at all, for any purpose. An investigation
of fallacies in historical scholarship may provide criteria by which some
of these deficiencies can be discovered and put right.

But if there are some ways in which a study of error can help his-
torical scholarship, there are others in which it can hurt. Popper's first
stage of knowledge — conjecture — in its earliest and most important
phases is not presently susceptible to rational analysis. There is no logic
of creative thought. Creativity makes its own rules. Genius transcends
them. The aboriginal act of inspiration remains utterly mysterious to
human understanding. We know when it happens, but not how or why.
It would be a very grave mistake to apply a logic for the testing of con-
jectures to conjecturing itself.

Equally important, though logic can distinguish error from truth
and truth from truism, it cannot distinguish a profound truth from a
petty one. A good many historical arguments are objectionable not
because they are fallacious but because they are banal, shallow, or
trivial. As a remedy for these failings, logic is impotent. Indeed, as I
collected material for this book, I quickly discovered that errors of the
sort I was looking for were most easily found in the work of the best and
brightest historians who are writing today. Many mindless monographs
call to mind Davy Crockett's critique of an effusion by Andrew Jackson —
"It don't even make good nonsense." There can never be a logic of grunts
and grimaces, nor a logic of the great clouds of conceptual confusion
which swirl around the heads of some historians. The thoughts of many
historians are neither logical nor illogical, but sublogical. To their work,
this book will be irrelevant.

Another qualification is also worth keeping in mind. Logical and
methodological techniques are not ends but means. It would be unfor-
tunate if historians were to become so obsessed by problems of how to
do their work that no work could ever get done. Abraham Kaplan was
warned against the "myth of methodology," the mistaken idea that "the
most serious difficulties which confront behavioral science are 'metho-
dological,' and that if only we hit upon the right methodology, progress
will be rapid and sure." This attitude is not merely unproductive, but
potentially destructive.

By pressing methodological norms too far [Kaplan writes] we may inhibit
bold and imaginative adventures of ideas. The irony is that methodology
itself may make for conformism — conformity to its own favored recon-

INTRODUCTION

xxi

structions. . . . And the push toward logical completeness may well make
for "premature closure" of scientific conceptions. The situation in science
is not unlike that in the arts: the critic with his standards discourages
daubers, but he also becomes the mainstay of the Academy, and art even-
tually passes by him. 8

No method exists independently of an object. None can be vindi-
cated except in its application; none can be proclaimed to the world as
The Method; and none is other than a useful tool, or more than an ap-
proximate tool. No historical method is in any sense an alternative to
heavy labor in historical sources. None can serve as a substitute for
creativity.

Conscious methodologies are not an indispensable prerequisite to
substantive success. Max Weber has written that

Methodology can only bring us reflective understanding of the means
which have demonstrated their value in practice by raising them to the
level of explicit consciousness; it is no more the precondition of fruitful
intellectual work than the knowledge of anatomy is a precondition for
"correct" walking. Indeed, just as a person who attempted to govern his
mode of walking continuously by anatomical knowledge would be in danger
of stumbling, so the professional scholar who attempted to determine the
aims of his own research extrinsically on the basis of methodological
reflections would be in danger of falling into the same difficulties. 9

But in historical scholarship, these are distant dangers. Most his-
torians are far removed from methodological obsessions — too far re-
moved, for the good of their discipline. Indeed, in a strict sense, academic
history today sometimes seems to be not a discipline at all, but a means
of teaching and writing without one. Among my professional brethren,
there is even a band of methodological Nullbruder, who flaunt their
intellectual poverty as if it were a badge of grace, and flourish all the
rusty instruments of ignorance in the face of every effort at reform.

The work of too many professional historians is diminished by an
antirational obsession — by an intense prejudice against method, logic,
and science. In their common speech, "scientism" has become a smear
word, and "scientific history" is a phrase which is used merely to con-
demn the infatuation of an earlier generation. In the process of this
reaction, historians have not merely severed their ties with the natural
sciences, but have also turned away from science in the larger sense of
a structured, ordered, controlled, empirical, rational discipline of thought.

History, it is said, is an inexact science. But in fact historians are

8. Abraham Kaplan, The Conduct of Inquiry (San Francisco, 1964), pp. 25-26.

9. Max Weber, "The Logic of the Cultural Sciences," in The Methodology of the Social
Sciences, trans. Edward Shils and Henry Finch (Glencoe, 111., 1949), p. 115.

xxii

INTRODUCTION

inexact scientists, who go blundering about their business without a
sufficient sense of purpose or procedure. They are failed scientists, who
have projected their failures to science itself. Nothing could be more
absurd, or more nearly antithetical to the progress of a potent discipline.

PART I

CHAPTER I

FALLACIES OF

QUESTION-FRAMING

Are we to be disgusted with science because it has not
fulfilled our hopes or redeemed its promises? And are
we, for this reason, to announce the "bankruptcy" of
science, as is so often and so flippantly done? But this is
rash and foolish; for we can hardly blame science just
because we have not asked the right questions.

Scissors-and-paste historians study periods; they collect
all the extant testimony about a certain limited group of
events, and hope in vain that something will come of it.
Scientific historians study problems: they ask questions,
and if they are good historians they ask questions which
they see their way to answering.

A moment's reflection should suffice to establish the simple proposition
that every historian, willy-nilly, must begin his research with a question.
Questions are the engines of intellect, the cerebral machines which con-
vert energy to motion, and curiosity to controlled inquiry. There can be
no thinking without questioning — no purposeful study of the past, nor
any serious planning for the future. Moreover, there can be no question-
ing in a sophisticated sense without hypothesizing, and no systematic
testing of hypotheses without the construction of hypothetical models
which can be put to the test.

Often, this intricate process is partly hidden from a historian, as
well as from his readers. Occasionally it is entirely invisible. But always
it exists. Without questions of some sort, a historian is condemned to
wander aimlessly through dark corridors of learning. Without questions

— Ernst Cassirer

— R. G. Collin gwood

4 INQUIRY

of the right sort, his empirical projects are consigned to failure before
they are fairly begun.

Specific forms of question-framing depend in a considerable degree
upon the kinds of answers which are sought. There are, of course, wide
variations in common practice. But there are also a few common denom-
inators of question-framing. These elemental aspects of questioning are
common to all historical inquiry, and indeed to empirical investigation
in every field. They are the business of this chapter. 1

It should be self-evident that some questions will yield empirical
answers and others will not. How does one distinguish the latter from
the former? This chapter will proceed first to an examination of ten
fallacies of empirical question-framing which have actually — and often —
occurred in historical scholarship. Ten more could easily be added. But
the following fallacies account for most of the erroneous questions I
have found. After a survey of these various forms of error, the chapter
will end with an attempt to articulate a few affirmative axioms.

The Baconian fallacy 2 consists in the idea that a historian
can operate without the aid of preconceived questions, hypotheses,
ideas, assumptions, theories, paradigms, postulates, prejudices, pre-
sumptions, or general presuppositions of any kind. He is supposed to go
a-wandering in the dark forest of the past, gathering facts like nuts and
berries, until he has enough to make a general truth. Then he is to store
up his general truths until he has the whole truth. This idea is doubly
deficient, for it commits a historian to the pursuit of an impossible object
by an impracticable method.

1. Nothing in this chapter is unique to historical inquiry. The reader will find close
parallels between practices discussed here and an analysis of question-framing in survey
research. Compare Stanley L. Payne, The Art of Asking Questions (Princeton, 1951).

2. This form of error takes its name from Francis Bacon's articulation of a method
which "derives its axioms from the senses and particulares, rising by a gradual and
unbroken ascent, so that it arrives at the most general axioms last of all" {Novum
Organon, bk. 1, xix). It should be noted that this is unfair to Bacon, and inaccurate as
an understanding of his thought. Bacon's larger work, of which the Novum Organon is
but a part, did not defend an induction as simple-minded as this, but rather a more
complex method of interdependent inquiry and research. Bacon was no more a Baconian
than Marx was a Marxian, or Plato a Platonist. See Benjamin Farrington, Francis
Bacon: Philosopher of Industrial Science (New York, 1949), chap. 6; and F. H. Ander-
son, The Philosophy of Francis Bacon (Chicago, 1948), p. 292, passim; and F. Smith
Fussner, The Historical Revolution (New York, 1962), chap. 10.

Though the name is objectionable in this respect, I have adopted it because it is
standard, on H. W. B. Joseph's assumption that "If it is useful to have a nomenclature
of fallacies, it is useful to have a standard nomenclature." (An Introduction to Logic
[London, 1906], p. 533.)

FALLACIES OF QUESTION-FRAMING 5

The impracticable method is a simple induction from the particular
to the general. It cannot work, because there is an infinity of particulars
in the past. Their truth value is an objective entity that exists indepen-
dently of an inquirer. But their particularity is separately denned by
each inquiry. If a fact is a true statement about past events, then there
is no practicable limit to the number of facts which are relevant to even
the smallest historical problem. "Truths are as plentiful as falsehoods,"
writes a distinguished logician, "since each falsehood admits of a negation
which is true. Scientific activity is not the indiscriminate amassing of
truths; science is selective and seeks the truths that count most." 3

The impossible object is a quest for the whole truth — a quest which
characteristically takes one of three forms. Occasionally, it consists in
an attempt to know everything about everything. Sometimes it seeks to
learn something about everything. Most often it is a search for every-
thing about something. None of these purposes is remotely realizable.
A historian can only hope to know something about something.

The most common everything-about-something school imagines
that historical science might be constructed on the same architectural
principles as the Pyramid of Khufu, with monographs stacked upon
thick square monographs in one vast granite pile, the whole massy
structure to be crowned some day with the gilded figure of a historio-
graphical Newton.

But a glance at the history of historical writing suggests that this
is not at all the way in which historiography develops. The monographs
do not commonly come first and the general interpretations second.
Instead some master architect — not master builder — draws a rough
sketch of a pyramid in the sand, and many laborers begin to hew their
stones to fit. Before many are made ready, the fashion suddenly changes
— pyramids are out; obelisks are in. Another master architect draws a
sketch in the sand, and the hewing and chipping starts all over again. A
few stones can be salvaged, but most have to be cut from scratch. As
Huizinga writes, "when the master builder comes, he will find most of
the stones you have laid ready for him unusable." 4

This does not mean, however, that relativists receive the last word.
There are many objective truths to be told about the past — great and
vital truths that are relevant and even urgent to the needs of mankind.
But there is no whole truth to be discovered by a simple method of in-
duction. Every true historical statement is an answer to a question which
a historian has asked. Not to The Question. Not to questions about
everything. But to questions about something.

3. Willard Van Orman Quine, Methods of Logic (New York, 1959), p. xi.

4. Johan Huizinga, Men and Ideas (New York, 1959), p. 20.

6 INQUIRY

There are signs that historians are slowly beginning to learn this
lesson. All the classic examples of the Baconian fallacy derive from the
work of an earlier generation. The best and clearest illustration is not
Bacon, or Ranke, but a distinguished French historian, Fustel de Cou-
langes (1830-1889). Fustel produced two great works — The Ancient
City and a monumental History of the Political Institutions of Ancient
France, in six erudite volumes. He was also a splendid teacher. One
day, his students responded to a lecture with an ovation. "Do not applaud
me," Fustel is said to have replied. "It is not I who speaks to you, but
history which speaks through my mouth."

Fustel asserted that a historian should bring no preconceived ideas
to his research — not even questions or working hypotheses:

Since be cannot know the cause beforehand, he should not be content
to study a specific category of facts; he should carefully observe all the
facts, all the institutions, all regulations public or private, all the customs
of domestic life, and particularly everything that relates to the possession of
land, [sic/] He should study all of these things with equally careful attention,
for he does not know beforehand from which side enlightenment will come
to him. This method is slow, but it is the only one which is sure. It is not
the method of the doctrinaire, but of the inquirer. 5

Fustel conceded the possibility of error, and, indeed, affirmed that
"one day of synthesis demands years of analysis. In these researches
which require so much patience and so much effort, so much prudence
and so much boldness, the opportunities for error are innumerable, and
none can hope to escape it." But that error derived, in Fustel's thinking,
not from the historian's bias, but from the gaps in his evidence. The
danger in all of this is well described by G. P. Gooch, who reports that
Fustel "regarded his results as independent of himself and felt criticism
as something like blasphemy." 6 Moreover, he imagined that he had
diminished the nationalist bias which marred the work of his many
chauvinistic colleagues. But he had merely disguised it. In his major
work, written immediately after the Franco-Prussian War, the main
thrust was a minimization of the significance of Teutonic influences
which other scholars had found in the development of French and
English institutions. 7

Historians today, several generations removed from Fustel de Cou-
langes, rarely commit this fallacy in so blatant a form. The relativists

5. Fustel de Coulanges, Histoire des Institutions Politiques de VAncienne France, 6 vols.
(Paris, 1890), 5.xiii.

6. Ibid., 1 : xiii, 145; G. P. Gooch, History and Historians in the Nineteenth Century,
new ed. (Boston, 1939), p. 202.

7. See generally Histoire des Institutions Politiques, I, II; and also The Origin of
Property in Land, trans. Margaret Ashley (London, 1926).

FALLACIES OF QUESTION-FRAMING

7

have seen to that — it was their most constructive accomplishment. But
old error still survives, deep in the dark recesses of every historian's
heart. Every now and then, he is apt to slip a little and allow the guilty
secret to escape, like the character in an English academic novel who
picked up the phone and said. "History speaking!"

Some belletristic historians occasionally yield to Dr. Johnson's
irritable opinion that "questioning is not a mode of conversation among
gentlemen." Samuel Eliot Morison, reviewing a book which took issue
with a Morisonian interpretation, distinguished between the "realm of
hypothesis" and "the realm of history," in the majestic spirit of Newton's
"Hypotheses non Fingo." 8

Other scholars hold with Kipling that "them that asks no questions,
isn't told a lie." G. R. Elton skirts this position, in an essay on historical
method.

Preconceived notions are a much greater danger to historical truth than
either deficiency of evidence or error in detail [he wrote]. . . . Sociologists
establish "models" which they test by supposedly empirical evidence. To
an historian this seems a very dangerous procedure: far too often the
model seems to dictate the selection of facts used to confirm it. . . . The
historian must certainly make one initial choice, of main area of study or
line of approach. But after that (if he is worth considering at all) he becomes
the servant of his evidence of which he will, or should, ask no specific
questions until he has absorbed what it says. 9

But one might revise Kipling's statement and assert with equal accuracy
and greater relevance that "them that asks no questions isn't told a
truth."

In contemporary historiography, there is a tendency not to reject
this statement in an abstract way but rather to accept it in principle
and to forget it in practice. There is an inherited antipathy to questions
and hypotheses and models, which is apt to run below the surface of a
historian's thought. The results are readily apparent in the conceptual
poverty of many historical monographs — a poverty to be explained not
by the stupidity of the authors, but rather by their habitual reluctance to
give sufficient attention to the organization of their inquiry, to the speci-
fication of their assumptions, and to the explication of their intentions.
Some younger historians, particularly an embattled group who call them-
selves the New Economic Historians — Douglass North, Robert Fogel,
Albert Fishlow, Peter Temin, Lance Davis, and many others — have

8. American Historical Review 67 (1961): 89.

9. G. R. Elton, The Practice of History (New York, 1967), pp. 36, 38n, 62. In other
parts of this work, the author adds qualifications. But his position, as I understand it,
comes close to that of Fustel, and violates the succinct advice of Lord Acton: "Study
problems, not periods."

8 INQUIRY

begun to revolutionize their special field partly by making their own
questions and hypotheses and models as carefully and consciously ex-
plicit as possible. Other historians have much to learn from their con-
structive example. "The wise man is not he who avoids hypotheses,"
a German scholar has written, "but he who asserts the most probable,
and who knows best how to estimate the degree of their probability." 10

The fallacy of many questions is a common form of error,
which has been variously defined as: ( 1 ) framing a question in such a
way that two or more questions are asked at once, and a single answer
is required; or (2) framing a question in such a way as to beg another
question; or (3) framing a question which makes a false presump-
tion; or (4) framing a complex question but demanding a simple
answer.

"Have you stopped beating your wife?" This, the classic textbook
example, presumes, of course, that you have already begun to do so — a
presumption which is not merely ungenerous but possibly mistaken.
Many wife-beating questions were deliberately concocted by that playful
monarch Charles II, who enjoyed assembling the learned gentlemen of
his Royal Society and asking them, with a sovereign contempt for logic
as well as fact, to explain "why a live fish placed in a full bowl of water
does not cause it to overflow, while a dead fish does cause it to overflow."
None of his scholars dared to fault a royal question. Instead, they in-
vented answers of magnificent absurdity as an act of homage to a man
who was himself a consistent living argument for republicanism.

Historians, and others who attempt to think historically, have often
committed the same error without intending it. A specimen is supplied
by a distinguished sociologist, who wrote, in the context of the history
of slavery in America: "There exists a major problem about American
slavery, one on which a reader of even the best American historians on
slavery will not be enlightened: indeed, if he limits his reading to his-
torians he will hardly know that a problem exists. Why was American
slavery the most awful the world has ever known?" 11 But was American
slavery "the most awful the world has ever known"? Glazer frames his
question in such a way as to beg another question.

Other examples fly thick and fast in the historiography of the Amer-
ican Reconstruction. Don E. Fehrenbacher, in an essay on the state of

10. Hans Vaihinger, The Philosophy of "As If," trans. C. K. Ogden (London, 1924),
p. 89.

11. Introduction by Nathan Glazer to Stanley N. Elkins, Slavery: A Problem in Ameri-
can Institutional and Intellectual Life, 2d ed. (New York, 1963), p. ix.

FALLACIES OF QUESTION-FRAMING

9

the literature, poses the following questions which he believes to be
"especially worth asking and answering."

1. "Was Reconstruction shamefully harsh or surprisingly lenient?"

2. "Was the presidential plan of reconstruction a sound one?"

3. "Could Lincoln have succeeded where Johnson failed?"

4. "Was the latter a miserable bungler or a heroic victim?"

5. "What were the primary motives of the Radical Republicans?"

6. "How bad were the carpetbag governments?"

7. "How well did the freedman meet his new responsibilities?"

8. "What part did terrorism play in the ultimate triumph of the Southern
'redeemers'?"

9. "When did racial segregation harden into its elaborate mold?" 12

The third question commits the fallacy of fictional questions, which
is discussed below. All others are examples of the fallacy of many ques-
tions, by any of the definitions listed above. Fehrenbacher's first question
assumes that Reconstruction was either "shamefully harsh" or "surpris-
ingly lenient," but maybe it was something else again. The second ques-
tion assumes that there was a single presidential plan of reconstruction,
which is doubtful. The fourth commits precisely the same sort of error
as the first; the fifth assumes that there were some clearly primary rad-
ical motives, and thereby encourages a simple motivational monism so
common in historical writing. The sixth, literally construed, assumes
that the carpetbag governments were bad in some degree; the seventh
assumes that freedmen in fact had new responsibilities, which were met
in some degree. The eighth assumes that the Redeemers did "ultimately"
triumph. The ninth assumes that racial segregation did at some point
in time harden into an elaborate mold, but maybe that institution has
been continuously in process of change.

There are other complaints to be entered against Fehrenbacher's
questions. They are mostly metaphysical questions and counterquestions,
and they are marred by the heavy-handed moralizing which has so
seriously diminished Reconstruction historiography. But the prior com-
plaint in this chapter is that all of them commit the fallacy of many
questions.

The fallacy of false dichotomous questions is a special form
of the fallacy of many questions, which deserves to be singled out for
special condemnation. It arises from the abuse of an exceedingly dan-
gerous conceptual device. Dichotomy is a division into two parts. If it

12. Don E. Fehrenbacher, "Division and Reunion," in John Higham, ed., The Recon-
struction of American History (New York, 1962), p. 105.

10 INQUIRY

is properly drawn, the parts are mutually exclusive and collectively ex-
haustive, so that there is no overlap, no opening in the middle, and nothing
omitted at either end. These three requirements are very difficult to
satisfy in the organization of an empirical inquiry. It is rare that any
two historical terms can be so related, unless one of them is specifically
defined as the negation of the other. And even then, there is often trouble.
The law of the excluded middle may demand instant obedience in formal
logic, but in history it is as intricate in its applications as the internal
revenue code. Dichotomy is used incorrectly when a question is con-
structed so that it demands a choice between two answers which are in
fact not exclusive or not exhaustive. But it is used often by historians in
this improper way. Indeed, a little industry has been organized around
it: the manufacture of the "problems" series of pamphlets for pedagogical
purposes. These works conventionally begin with a false dichotomous
question, allegedly designed to "stimulate" thought. The question takes
the form of "Basil of Byzantium: Rat or Fink?" Maybe Basil was the
very model of a modern ratfink. Maybe that Byzantine character was
neither a rat nor a fink, but something vastly more intricate, or something
altogether different. But swarms of suffering undergraduates are asked to
study a set of pedantical essays, half of which are exaggerated arguments
for the rattiness of Basil and the other half are overdrawn portraits of
Basil as a fink. The disgusted undergraduate is expected to make a choice
between these unappetizing alternatives, or perhaps to combine them in
some ingenious paradoxical contrivance of his own invention, which
falsifies both his understanding and the problem itself.

The following examples are the actual titles of works which have
been edited by reputable professional historians and issued by respec-
table publishers such as Holt, Rinehart; Prentice-Hall; Houghton Mif-
flin; Random House; and D. C. Heath:

Napoleon III: Enlightened Statesman or Proto-Fascist?

The Causes of the War of 1812: National Honor or National
Interest?

The Abolitionists: Reformers or Fanatics?
Jacksonian Democracy: Myth or Reality?
Plato: Totalitarian or Democrat?
The Dred Scott Decision: Law or Politics?

The Removal of the Cherokee Nation — Manifest Destiny or
National Dishonor?

John D. Rockefeller — Robber Baron or Industrial Statesman?
The Robber Barons — Pirates or Pioneers?

Huey P. Long — Southern Demagogue or American Democrat?
The New Deal — Revolution or Evolution?

FALLACIES OF QUESTION-FRAMING 11

Industry-Wide Collective Bargaining — Promise or Menace?
Ancient Science — Metaphysical or Observational?
Feudalism — Cause or Cure of Anarchy?
The Medieval Mind — Faith or Reason?

The Parliament of Edward I — Royal Court or Representative
Legislature?

Renaissance Man — Medieval or Modern?

Martin Luther — Reformer or Revolutionary?

The Absolutism of Louis XIV — The End of Anarchy or the Be-
ginning of Tyranny?

The Scientific Revolution — Factual or Metaphysical?

The Industrial Revolution in England — Blessing or Curse to the
Working Man?

The Fall of the Russian Monarchy — Inherent Failure or Planned
Revolution?

The Origins of Nazi Germany — German History or Charismatic
Leadership?

What Is History — Fact or Fancy?

Many of these questions are unsatisfactory in several ways at once.
Some are grossly anachronistic; others encourage simple-minded moral-
izing. Most are very shallow. But all are structurally deficient in that
they suggest a false dichotomy between two terms that are neither
mutually exclusive nor collectively exhaustive. They are also imprecise,
both in the dichotomous terms and in the troublesome connective "or,"
which might mean "either X or Y but not both" (like the Latin aut),
or "either X or Y or both" (like the Latin vel), or "either A" or Y or
both, or neither." This ambiguity is not often clarified in context, and
the reader receives no clear indication of what he is being asked to
accomplish.

The "problems" that appear in these pamphlets are not merely a
result of faulty pedagogical practice. Many of these titles reflect a false
dichotomy which is deeply embedded in scholarly literature on the
subject at hand. They are illustrations not only of the way in which
many historians teach but also of the way in which they conceptualize
and carry on their own research.

What can a student do, in the face of a false dichotomy? He can
try several stratagems. First, he might attempt to show that the dichot-
omous terms can coexist. Second, he might demonstrate that there is a
third possibility. Third, he might repudiate one or the other or both
alternatives. All of these devices will work, in a limited way. But all
of them will have the effect of shackling the student's answer to the
fallacious conceptualization he is attempting to correct. The most satis-

12 INQUIRY

factory response, I think, is to indicate the structural deficiencies in the
question-framing and to revise the inquiry on that level, by the intro-
duction of a more refined and more open question, which can be
flexibly adjusted as the analysis proceeds.

The problem of an exclusive choice between nonexclusive alter-
natives is often confronted in declarative as well as interrogative sen-
tences. The motto of the Prince of Orange was Non rapui sed recepi,
which means, "I didn't steal; I received." But maybe he was a receiver
of stolen goods, even if they were stolen in a glorious cause. In
this question, as in so many others, one can only endorse the sensible
observation of Reuben Abel: "The continuum in which we live is
not the kind of place in which middles can be unambiguously ex-
cluded." 13

The fallacy of metaphysical questions is an attempt to resolve
a nonempirical problem by empirical means. In its most common contem-
porary form, this fallacy consists in the framing of a question which
cannot be resolved before the researcher settles some central metaphysical
problem such as "What is the nature of things?" or "What is the inner
secret of reality?" And these are questions which will not be resolved
before the oceans freeze over.

A prime example is the problem which is eternally popular among
Civil War historians: "Was the War inevitable?" A scholar who carries
this question to the archives can illustrate his answer by reference to
historical events; he can add persuasive power to his metaphysical prop-
osition by the appearance of factual solidity. But he can no more hope
to resolve the issue of inevitability by empirical research than he can
hope to determine by modern methods of quantification the number of
angels which might be made to perch upon the head of a proverbial
pin.

If Thomas J. Pressly is correct, in his study of historical interpreta-
tions of the Civil War, the problem of inevitability has been the central
historiographical problem. 14 A few historians have repudiated it, such
as Kenneth Stampp, who tried to turn away from "the fruitless and
impossible task of proving or disproving that the Civil War was inevi-
table." 15 In the same spirit, Arthur Schlesinger, Jr., has written that "the
problem of the inevitability of the Civil War, of course, is in its essence

13. "Pragmatism and the Outlook of Modern Science," Philosophy and Phenomenological
Research 27 (1966): 50.

14. Thomas J. Pressly, Americans Interpret Their Civil War (Princeton, 1954).

15. Kenneth M. Stampp, And the War Came, 2d ed. (Chicago, 1964), p. vii.

FALLACIES OF QUESTION-FRAMING 13

a problem devoid of meaning." 10 And Pieter Geyl has declared that
"The question of evitable or inevitable is one on which, it seems to me,
the historian can never form any but an ambivalent opinion." 17

But all of these scholars were themselves unable to keep clear
of the problem they condemned. Their works are refutations of the
argument that the Civil War was a "needless war," which was precipi-
tated by a "blundering generation" of American political leaders. 1 "
And there are discouraging signs that the problem of inevitability is
still obsessive in the consciousness of the coming generation of young
American historians. 19

"As a historian," E. H. Carr writes, "I am perfectly prepared to do
without 'inevitable,' 'unavoidable.' 'inescapable,' and even 'ineluctable.'
Life will be drabber. But let us leave them to poets and meta-
physicians." Historians might also turn their backs upon all aspects of
the metaphysical problems raised by determinism versus voluntarism,
the comparative reality of individuals and groups, materialism versus
idealism, and all manner of other monisms and dualisms. The progress
of an empirical science of history squarely depends upon a sense of the
possible. The working historian, in my opinion, is well advised to deal
with these dilemmas by a method of indifference. 20

Some historians of a humanist bent will protest that all historical
problems are metaphysical problems. This is humbug. It can be argued
that all historical problems can be made into metaphysical problems if

16. Arthur Schlesinger, Jr., "The Causes of the Civil War: A Note on Historical
Sentimentalism," Partisan Review 16 (1949): 980.

17. Pieter Geyl, "The American Civil War and the Problem of Inevitability," Debates
with Historians (New York, 1958), p. 263.

18. Stampp, p. v; Schlesinger, pp. 970-81; Geyl, pp. 244-63. For two other thoughtful
critiques of the problem, see Lee Benson and Cushing Strout, "Causation and the
American Civil War," History and Theory 1 (1960): 163-85.

19. See, e.g., Eugene D. Genovese, The Political Economy of Slavery (New York, 1965),
pp. 4-10, passim.

20. E. H. Carr, What Is History? (New York, 1962), p. 126. Carr believes that the in-
evitability problem "attaches itself almost exclusively to contemporary history. Last term
at Cambridge," he writes, "I saw a talk to some society advertised under the title: 'Was
the Russian Revolution Inevitable?' I am sure it was intended as a perfectly serious talk.
But if you had seen a talk advertised on 'Were the Wars of the Roses Inevitable?' you
would at once have suspected some joke. The historian writes of the Norman conquest
or the American war of independence as if what happened was in fact bound to happen,
and as if it was his business simply to explain what happened and why" (ibid.).

But this, unfortunately, is not true, at least as far as the American Revolution is
concerned. David Hawke writes, "Nothing intrigues colonial historians more than the
question of why the American Revolution occurred. Was it a repressible conflict or not?"
(The Colonial Experience [Indianapolis, 19661, p. 517.) I think it probable that his-
torians debate the "repressibility" of every important happening which they are unable
to accept unanimously as a Good Thing.

14 INQUIRY

one wishes to do so. But there are many historical problems of primary
importance to all inquirers, whatever their opinions may be, which are
clearly not metaphysical. "How many people voted in the election of
1840?" "How did the price of cotton change in the 1850s?" "What did
the Halfway Covenant mean to the men who made it?" "Was Franklin
Roosevelt more interventionist than a majority of the American people
in 1940, or less so?" Nonmetaphysical questions can be exceedingly
complex and sophisticated. "How and when did habits of authority
develop and decline in English and American politics?" "How did the
personality patterns of Negro slaves change during the period of their
enslavement?"

These are urgent questions, and they are empirical questions, which
can be put to the test. The reader will note that none of them are "why"
questions. In my opinion — and I may be a minority of one — that favorite
adverb of historians should be consigned to the semantical rubbish heap.
A "why" question tends to become a metaphysical question. It is also
an imprecise question, for the adverb "why" is slippery and difficult to
define. Sometimes it seeks a cause, sometimes a motive, sometimes a
reason, sometimes a description, sometimes a process, sometimes a
purpose, sometimes a justification. A "why" question lacks direction
and clarity; it dissipates a historian's energies and interests. "Why did
the Civil War happen?" "Why was Lincoln shot?" A working historian
receives no clear signals from these woolly interrogatories as to which
way to proceed, how to begin, what kinds of evidence will answer the
problem, and indeed what kind of problem is raised. There are many
more practicable adverbs — who, when, where, what, how — which are
more specific and more satisfactory. Questions of this sort can be resolved
empirically, and from them a skilled historian can construct a project
with much greater sophistication, relevance, accuracy, precision, and
utility, instead of wasting his time with metaphysical dilemmas raised
by his profound "why" questions, which have often turned out to be
about as deep as the River Platte.

It is improbable that this will happen, among historians, in the
foreseeable future. "Why" questions are rooted in the literature and
institutionalized in the graduate schools, and for most of my colleagues
a historical discipline without them is as strange as a system of non-
Euclidean geometry. But it is already beginning to happen, in a quiet
way, in monographs such as Thomas Barrow's Trade and Empire: The
British Customs Service in Colonial America, 1 660-1775 (Cambridge,
1967), a fine book which orthodox academic reviewers have utterly
failed to understand. A young, able, but very old-fashioned scholar,
Benjamin Labaree, complains in his notice of the book that "Barrow

FALLACIES OF QUESTION-FRAMING 15

does not explain with full satisfaction why the Americans evaded the
Acts of Trade, but he does marshal a wealth of material to show how
they managed to disobey them." 21 One wonders what kind of "why"
answer would have satisfied Labaree. The hand of God? The dialectic?
Some mighty dynamic of materialism? Maybe some American merchants
helped to smash up the mercantile system for the same reason that
Kirillov destroyed himself in Dostoevsky's The Possessed — to demon-
strate that they were free. These questions are not for a historian, who
can only measure the motives and purposes that are part of the act
itself. He can never hope to find the inner secret, maybe because it does
not exist.

Be that as it may, Barrow's book is a straw in the wind. His work
suggests that there is a fair and steady offshore breeze which is blowing
historians clear of the rocks and shoals of metaphysics, though some
seem determined (an ambiguity is intended) to know the excitement
of a shipwreck, which is the only kind of metaphysical finality that is
open to them.

A rigorous attempt to purge history of metaphysics will, in truth,
serve to narrow historical inquiry. To those who protest that the result
would be a little too narrow, one might repeat the words of Nelson
Goodman: "You may decry some of these scruples and protest that there
are more things in heaven and earth than are dreamt of in my philosophy.
I am concerned, rather, that there should not be more things dreamt
of in my philosophy than there are in heaven or earth." 22

The fallacy of fictional questions is an ancient form of error,
which has recently been elevated Lito an explicit method and proclaimed
before the world as a whole new thing in historical inquiry. It consists
in an attempt to demonstrate by an empirical method what might have
happened in history, as if in fact it actually had: the sort of thing that
Philip Guedalla and others did in a lively work called //, or History
Rewritten (New York, 1931), in which they ruminated upon the might-

21. Journal of American History 54 (1968): 876.

22. Nelson Goodman, Fact, Fiction, and Forecast (Cambridge, 1955), p. 39. Some his-
torians have argued that the direction of historiographical change is the opposite of
that which I have suggested. Miss C. V. Wedgwood writes, "The older historians con-
centrated more on narrative than on analysis, on the How rather than the Why of
history. But now, for several generations, Why has been regarded as a more important
question than How." Truth and Opinion (New York, 1960), p. 14. This, I think,
is mistaken. Perhaps the change has been from the implicit why of older narrative
historians to the explicit Why of the last generation of monographers, to the controlled
who, how, where, when, and what of historians who are presently beginning to publish.

16 INQUIRY

have-beens, if Booth's bullet had missed, or Drouet's cart had stuck, or
Napoleon had escaped to America, or the Moors had won in Spain, or
Lee had won at Gettysburg, or Byron had become King of Greece.

There is nothing necessarily fallacious in fictional constructs, as
long as they are properly recognized for what they are and are clearly
distinguished from empirical problems. All novels are organized around
an idea of what might have happened — some very great truths have
been taught to the world in this disguise. Fictional questions can also
be heuristically useful to historians, somewhat in the manner of meta-
phors and analogies, for the ideas and inferences which they help to
suggest. But they prove nothing and can never be proved by an empirical
method. All historical "evidence" for what might have happened if
Booth had missed his mark is necessarily taken from the world in which
he hit it. There is no way to escape this fundamental fact. 23

Nevertheless, fictional questions have become fashionable among
a group of young economic historians who call themselves "Cliometri-
cians" and are doing many good and revolutionary things to their dis-
cipline, in the way of an extension of empirical inquiry, a refinement
of conceptualization, and a general integration of history and theory.
Economic theory is theoretical in the narrow sense of the word. It deals
in "if, then" propositions, unlike much social "theory," which is com-
monly more paradigmatic in its nature. It is understandable that eco-
nomic historians, more than others, are tempted by the seductive might-
have-beens of "as if questions. A few have even attempted to combine
"fictional," "counterf actual," "conditional," or "as if questions (as they
are variously called) with techniques of empirical quantification. The
results are not merely false but absurd, for to quantify the conditional
is to square the circle. It is simply impossible for a singular statement
to be both counterfactual and factual at the same time.

The classic example is a controversial monograph by Robert W.
Fogel, Railroads and Economic Growth: Essays in Econometric History
(Baltimore, 1964), in which the author tries to prove that railroads
were not "indispensable" to American economic growth in the nine-
teenth century by demonstrating how the economy might have functioned
if railroads had not existed. Fogel measures "primary effects" of the
transportation system in terms of the costs of haulage by railroads, as

23. There is an immense philosophical literature on the subject of fictional questions,
or "counterfactual conditionals." The best discussion I have seen is in Nelson Goodman,
Fact, Fiction, & Forecast, 2d ed. (Indianapolis, 1965), pp. 13-34, in which the author
emphatically declares that "a counterfactual by its nature can never be subjected to any
direct empirical test by realizing its antecedent" (p. 4). This is the only point which I
mean to argue here.

FALLACIES OF QUESTION-FRAMING 17

against haulage exclusively by turnpikes and canals, at actual nine-
teenth-century prices. He also estimates "secondary effects" — changes
in spatial distribution, generation of demand for manufactures, etc.
From these figures, he calculates that the "social saving" derived from
the use of railroads was comparatively small, as a proportion of the
gross national product. And he concludes that railroads were in fact
"dispensable" to economic development in nineteenth-century Amer-
ica. 24

Fogel's book might have been called //, or History Derailed. His-
tory is run off the rails in more senses than one. There is much to be
salvaged from the wreckage — a zeal for quantification of historical
problems, a determination to make questions and assumptions explicit,
and an impressive conceptual sophistication. But Fogel's inquiry is flawed
by three fatal inconsistencies: First, his evidence of a transportation net
which might have operated in the absence of railroads is necessarily
derived from a world in which railroads were present. Fogel tries to
allow for this bias in his material, but only by introducing other fictional
constructs which assume what he promises to prove by empirical inquiry.
The cost of haulage by canal boats is merely one of many imponder-
ables. Fogel bases his estimates upon a situation where canals and rail-
roads coexisted. It is possible that competition between the two served
in some measure to reduce rates for canal travel, but it is equally possible
that if canals had been the mainstay of the transportation system, they
would have been more efficient in their operations and a spur to tech-
nological innovation, which might have reduced rates. Who can say
which? And yet, the question is critical to Fogel's thesis. Moreover, the
secondary, tertiary, etc., effects of his canals upon industry, immigration,
national psychology, national politics, and such specific happenings as
the Civil War cannot even be guessed, much less proved by empirical
research.

Second, there is another serious flaw in Fogel's logic. He believes
that "to establish the proposition that railroads substantially altered the
course of economic growth, one must do more than provide information
on the services of railroads. It must also be shown that substitutes for
railroads could not (or would not) have performed essentially the same
role" (p. 207). But this confuses two separate questions. It is one thing
to ask, "Did railroads alter the process of American economic growth?"

24. Edward Kirkland, an economic historian of the old school, complains that there is a
kind of doublethink in the work of Fogel and his friends. For the former, a social saving
of 5 percent of GNP appears small. But for Douglass North, "another card-carrying
Cliometrician, 'a social saving of 5 percent is a very substantial saving.' " American His-
torical Review 72 (1967): 1494.

18 INQUIRY

And quite another to ask, "Did railroads alter the process of American
economic growth in a way that only railroads were able to do?" The
first question is empirically verifiable. The second is not. And the first
question does not require an answer to the second. It can, for example,
be conclusively proved that Thomas Jefferson, as president, was the
agent of certain great and grave events, such as the purchase of
Louisiana. But nobody will ever know if he was "indispensable" to that
result, nor is the problem of his indispensability necessarily implied by
the fact of his agency. No amount of empirical research will ever suffice
to prove that Timothy Pickering, had he by some horrible twist of fate
been elevated to the presidential chair, would or would not have done
precisely what Jefferson did. His perverse opinions on Louisiana are
well known, but the opinions which he might have held in different
circumstances are utterly unknowable, and irrelevant to a proper his-
torical inquiry. And in precisely the same fashion, nobody will ever
know what miracles might have been wrought by Fogel's counterfactual
canal boats, which are not more mythical (if a little more plausible)
than similar vessels which some enthusiasts have spied plying the alleged
waterways of the planet Mars.

Third, the question of the "indispensability" of railroads is com-
parable to the problem of the "inevitability" of the Civil War. Fogel is
leading his Cliometrical colleagues down the methodological rathole
of the metaphysical question. His work is a forward step, in its explicit-
ness, sophistication, and attempt at quantification. But it is also a step
backward in its return to ancient metaphysical conundrums which have
distracted many generations of historians. This aspect of the New Eco-
nomic History is not new at all, but ancient, and even anachronistic.

Fogel's counterfactual canals have created an uproar among eco-
nomic historians. An immense controversial literature has appeared in
their journals. 25 Some of Fogel's colleagues have tried to salvage some-

25. Louis M. Hacker, "The New Revolution in Economic History: A Review Article
Based on Railroads and Economic Growth: Essays in Econometric History, by Robert
William Fogel," Explorations in Entrepreneurial History, 2d ser. 3 (1966): 159-75;
Lance Davis, "Professor Fogel and the New Economic History," Economic History Re-
view, 2d ser. 19 (1966): 657-63; R. W. Fogel, "The New Economic History: Its Find-
ings and Methods," ibid., 624-56; R. W. Fogel, "The Specification Problem in Economic
History," Journal of Economic History 27 (1967): 283-308; Alexander Gerschenkron,
"The Discipline and I," Journal of Economic History 27 (1967): 443-58; Meghnad
Desai, "Some Issues in Econometric History," Economic History Review, 2d ser. 21
(1968): 1-16; E. H. Hunt, "The New Economic History: Professor Fogel's Study of
American Railroads," History 53 (1968): 3-18; G. R. Hawke, "Mr. Hunt's Study of
the Fogel Thesis: A Comment," ibid., 18-23.

Fritz Redlich, in "New and Traditional Approaches to Economic History," Journal
of Economic History 25 (1965): 484-85, introduces to the controversy the work of

FALLACIES OF QUESTION-FRAMING 19

thing from his work by arguing that counterfactuals can be used correctly
"for the elucidation of relatively short-term changes, preferably in situa-
tions in which the political factor may be largely neglected. Once the
period under review lengthens, the number of unconsidered and noncon-
siderable factors that bear upon the outcome increases fast and the signi-
ficance of the results diminishes faster." 26 One wonders what this scholar
means by the "political factor." Maybe he is thinking of something
like the "human factor," in which case he would rule out any historical
problem involving those irritating and unpredictable things called people.
But be that as it may, his attempt at compromise will not work for other
reasons. The only difference between long-run and short-run counter-
factuals is that the absurdity of the former is more glaringly apparent.
It is not more extreme.

Fogel has himself replied to critics of counterfactualizing with the
argument that everybody does it, that the alternative to an open counter-
factual model is a concealed one." But this is surely a mistake, for rea-
sons discussed in the case of Jefferson and Louisiana, above. There is,
I think, an increasing body of historical literature which is noncausal
in its nature (in any meaningful or common sense of causality), and
there is some which has refined the problem of causality so as to exclude
problems such as inevitability and indispensability, by working closely
from the assumption that things happened merely in the way that they
happened and not in any other way. This is not to affirm a determinism,
nor to deny that men make choices, but merely to short-circuit the
problem and to get on to others that we can handle. It is always possible,
of course, to convert any historical problem into a nonhistorical one,
but why should a scholar go out of his way to make a difficult problem
impossible? History is tough enough, as it is — as it actually is.

It is true that many other historians besides Fogel have resorted
to explicit or implicit counterfactual models. But it is not true that they
must or should do so, for the same difficulties which developed in Fogel's

Hans Vaihinger — mistakenly, I think. He argues that Fogel's counterfactuals are "fig-
ments" in Vaihinger's terms. But they fit more closely into Vaihinger's category called
"fictions," viz., "imaginary cases." (Vaihinger, The Philosophy of "As If," p. 26.)
Vaihinger clearly distinguishes between "fictions" and "hypotheses." He argues that "the
latter are assumptions which are probable; assumptions, the truth of which can be proved
by further experience. They are therefore verifiable. Fictions are never verifiable; for
they are hypotheses which are known to be false [sic], but which are employed because
of their utility" (p. xlii). That is the point which is held for, here. Vaihinger observes
that hypotheses are true or false, but fictions are merely expedient or inexpedient.

26. Alexander Gerschenkron, "The Discipline and I," Journal of Economic History 17
(1967): 457. See also Sidney Hook, The Hero in History (Boston, 1943), chap. 7.

27. Fogel, "The Specification Problem," p. 285.

20 INQUIRY

work appear in all other cases. Consider, for example, the work of an
able political historian, Eric McKitrick. In an elaborate and sophisticated
study of Andrew Johnson and the American Reconstruction, which
was written in the years of the Eisenhower "consensus," McKitrick
wondered if "an imaginary peace-making" might have been arranged
between two influential Americans — Wade Hampton, a supposititious
spokesman for the South Carolina "establishment," and John Andrew,
allegedly of the Massachusetts "establishment." McKitrick, who appears
to believe that "establishments" are good things, and that all good
things, which are mostly conservative things, must be done through them,
has no difficulty demonstrating to his own satisfaction that such a
peace-making was possible — thus reflecting upon the statesmanship
of President Andrew Johnson, who lacked the style which McKitrick
respects. 28

It might be demonstrated that the author misinterprets the position
of Wade Hampton, which was far more intransigent than McKitrick is
prepared to allow. 29 But a more serious flaw consists in the attempt at
empirical reiflcation of an imaginary event, which exists only in the
mind of the author. The result is merely sentimental claptrap, which
calls to mind a phrase from Eliot:

Footfalls echo in the memory
Down the passage which we did not take
Towards the door we never opened
Into the rose-garden 30

McKitrick's imaginary peacemaking brings to mind imaginary
war-making, which seems almost as firmly established in the world today
as war itself. German soldiers and statesmen, before they made war
upon their various neighbors, formed the habit of conducting elaborate
kriegspiele on blackboards and sandtables, which must have been as much
fun as the real thing. On the eve of World War II, German leaders
played a war game among themselves, in which they demonstrated to
their own satisfaction that England could not and would not intervene
in Poland's interest. Today this grim gamesmanship is much in fashion
among semiacademic nuclear strategists at the Rand Corporation, and
the Hudson Institute, and, I am told, in the Red Army Historical Section
as well. Let us hope that they will not make a similar mistake, and put
their trust in a method which is not merely delusive to themselves but
also exceedingly dangerous to others.

28. EricL. McKitrick, Andrew Johnson and Reconstruction (Chicago, 1960), pp. 214-50.

29. Joel Williamson, After Slavery: The Negro in South Carolina During Reconstruction,
1861-1877 (Chapel Hill, 1967), pp. 276, 406-12.

30. T. S. Eliot, "Burnt Norton," Collected Poems, 1909-1962 (New York, 1963), p. 175.

FALLACIES OF QUESTION-FRAMING 21

If by some miracle of rationality the Cold War were brought to
an end, the Kriegspielers of the world would be faced with the dark
prospect of occupational obsolescence. In that desperate predicament,
Cold Warriors might beat their conceptual swords into historiographical
plowshares and begin to dig up the past. An inkling of the possible
results appears in "An Attempt to Simulate the Outbreak of World War
I," by Charles F. and Margaret G. Hermann. 31 The serious student of
that event may be interested to learn that according to the simulation,
England considered "the initiation of war on Germany while the advan-
tage appeared on her side" (presumably by sending the Old Contempti-
bles hopping over the low countries on pogo sticks). Moreover, the
Austro-Hungarian decision maker in the game suddenly "revealed
pacifistic tendencies and readily accepted the objections to his nation's
militaristic actions" (the nature of that extraordinary conversion ex-
perience is not explained). The authors conclude with a Scotch verdict
on their own efforts. "Until more validation exercises are conducted,"
they believe, "it is premature to accept or reject simulation as an impor-
tant new tool for studying political phenomena" (p. 416).

That judgment is safe, and maybe sound (I believe otherwise). But
in the meantime, Fogel and McKitrick, and the counterfactualists, and
the Kriegspielers, might profit from the advice of those learned, if unlov-
able logicians, Tweedledum and Tweedledee:

"I know what you're thinking about," said Tweedledum; "but it isn't so,
nohow."

"Contrariwise," continued Tweedledee, "if it was so it might be; and if
it were so, it would be; but as it isn't, it ain't. That's logic." 32

The fallacy of semantical questions consists in an attempt to
resolve, by empirical investigation of an object, a semantical question
about the name by which that object is called, thereby confusing actual
happenings with verbal descriptions of actual happenings. The problem
is a difficult one, much complicated by developments in an important
new discipline (linguistics), and a revolution in an old one (philosophy),
both of which have spawned a vast literature of sophisticated and useful
scholarship. All historical questions are semantical in some degree, in
that they are attempts to establish intelligible relationships between the
signs and symbols of our language on the one hand and the evidence of
our past on the other.

But some questions which historians have asked are merely seman-

31. The American Political Science Review 61 (1967): 400-16.

32. The Complete Works of Lewis Carroll, Modern Library ed. (New York, n.d.), p. 181.

22

INQUIRY

tical, which is to say that they are sterile disputes about word usage and
not about the past happenings to which the words are supposed to refer.
Semantical questions are "term questions," in the fashion of the origi-
nal term question, which was not the least of many miseries inflicted by
well-meaning Western missionaries upon the suffering souls of China.
But in fairness to the missionaries, it should be said that the Chinese seem
to be culturally predisposed to term questions. Witness the legendary
Ming emperor who dealt with a dangerous river by a sort of semantical
flood control project. Instead of building dikes and dams, he changed its
name from "The Wild One" to "The Peaceful One." 33 So common have
such "rectifications of names" been in that country, from Confucius
to Chou En-lai, that this form of error might be called the Chinese
fallacy.

Semantical questions are deeply embedded in our own culture as
well. A classical example is the conversation between Alice and the
White Knight in Through the Looking Glass:

"You are sad," the Knight said in an anxious tone: "let me sing you
a song to comfort you. . . . The name of the song is called 'Haddocks' Eyes' "

"Oh, that's the name of the song, is it?" Alice said, trying to feel in-
terested.

"No, you don't understand," the Knight said, looking a little vexed.
'That's what the name is called. The name really is 'The Aged Aged Man.' "

"Then I ought to have said 'That's what the song is called!" Alice cor-
rected herself.

"No, you oughtn't: that's quite another thing! The song is called
'Ways and Means': but that's only what it's called, you know!"

"Well, what is the song, then?" said Alice, who was by this time com-
pletely bewildered.

"I was coming to that," the Knight said. "The song really is 'A-sitting
On A Gate': and the tune's my own invention." 34

A historiographical example of the fallacy of semantical questions
presently appears in a prolonged dispute among American colonial his-
torians over the question, "Was the political structure of seventeenth-
century America 'democratic' or 'aristocratic'?"

Consider two articles on the subject, one by Mrs. B. Katherine
Brown, and another by Professor Roy Lokken. 35 Both quote a statement
by John Cotton, addressed to a Puritan Lord in England who was con-

33. Hans Konigsburger, Love and Hate in China (New York, 1966), p. 9.

34. The Complete Works of Lewis Carroll, p. 244.

35. B. Katherine Brown, "A Note on the Puritan Concept of Aristocracy," Mississippi
Valley Historical Review 41 (1954): 105-12; Roy N. Lokken, "The Concept of De-
mocracy in Colonial Political Thought," William and Mary Quarterly, 3d ser. 16 (1959);
568-80.

FALLACIES OF QUESTION-FRAMING 23

cerned about the alleged absence of Orders and Distinctions in the New
World. Cotton wrote:

Nor neede your Lordship feare (which yet I speake with submission to
your Lordships better judgment) that this corse will lay such a foundation,
as nothing but a mere democracy can be built upon it. Bodine confesseth,
that though it be a status popularis, where a people choose their owne
governors; yet the government is not a democracy, if it be administered,
not by the people, but by the governors, whether one (for then it is a
monarchy, though elective) or by many, for then (as you know) it is
aristocracy. 38

Cotton and John Winthrop and others called their polity a "Mixt
Aristocratie," but Mrs. Brown argues that given their definitions "Puritan
'Mixt Aristocratie' is much nearer our notion of democracy than our
current meaning of aristocracy" (p. 111).

Lokken cites the same evidence to sustain a contrary proposition,
that Cotton was hostile to "unmixed democracy," that there was nothing
in his thought like a nineteenth- or twentieth-century sense of democracy,
for "in the political ethos of the seventeenth and eighteenth century
English world, representation did not involve direct responsibility to the
electorate" (pp. 577-78).

This exchange has gone round and round for ten years and gives no
sign of playing out. Some useful research has resulted, but in these ar-
ticles of Mrs. Brown and Mr. Lokken, it has degenerated into a seman-
tical question. What they are really concerned about, I think, is not "What
was the political structure of seventeenth-century America?" but only
secondarily this, and primarily "What name shall we give it?" And the
dispute which has developed from that question closely resembles an
argument over the question "Is a zebra a white animal with black stripes
or a black animal with white stripes?"

A more intelligent and useful approach to the problem is surely
possible. Suppose that all evidence of seventeenth-century New England
were totally destroyed, except John Cotton's replies to Lord Say and
Lord Brooke. From that source alone, much could be discovered about
the polity of Massachusetts Bay, most of which Mrs. Brown and Mr.
Lokken have ignored. Cotton's correspondence, both as secondary evi-
dence of the government of the Bay and as primary evidence of the
principles and prejudices of one inhabitant, shows much that has devel-

36. "Copy of a Letter from Mr. Cotton to Lord Say and Seal in the Year 1636," in
Thomas Hutchinson, The History of the Colony and Providence of Massachusetts-Bay,
ed. Lawrence Shaw Mayo, 3 vols. (Cambridge, Mass., 1936) 1: 416.

24 INQUIRY

oped and ramified in the next three centuries of American history and
much that has disappeared as well.

There is profound sense of social ranks: "Two distinct ranks we
willingly acknowledge, from the Light of Nature and scripture; the one
of them called Princes, or Nobles, or Elders (amongst whom gentlemen
have their place) the other the people. Hereditary dignity or honours we
willingly allow to the former ..." (p. 410). But there also is, in Cotton's
distinction, a repudiation of "hereditary authority and power" in formal
institutional arrangements (p. 412) which exceeds, I believe, the pro-
fessions and practice of any European society in the year 1636.

There is much else besides in this extraordinarily fertile document
which alone is sufficient to sustain a monograph (and probably will,
someday). In the meantime, the controversy might stand as a byword and
an example to historians who are tempted to go rummaging through
dusty records for Haddocks' eyes and Aged Aged Men and other seman-
tical fancies.

George Boas's observation is perhaps relevant: "The song of the
barnyard cock," he writes, "is cock-a-doodle-do in English, cocorico in
French, and kikiericki in German and in Italian. Would anyone debate
on which is right or whether cocks in different countries sing different
songs?" 37

$*> The fallacy of declarative questions consists in confusing an
interrogative with a declarative statement. It violates a fundamental rule
of empirical question-framing, which requires that a question must have
an open end, which will allow a free and honest choice, with minimal bias
and maximal flexibility. If a historian goes to his sources with a simple
affirmative proposition that "X was the case," then he is predisposed to
prove it. He will probably be able to find "evidence" sufficient to illus-
trate his expectations, if not actually to sustain them. If, on the other
hand, he asks, "Was X or Y the case?" then he has an empirical advan-
tage, at least in some small degree. And if he asks "Was X or not-A', Y
or not-F, Z or not-Z ... the case?" and if he designs X, and Y, and Z
in such a way that his own preferences are neutralized, and if he leaves
the way open to refinements in the form of Xi, Y\, Zi, and if he allows
for still other unexpected possibilities, then the probability of empirical
accuracy is still further enhanced.

A historian, like any other researcher, has a vested interest in answer-
ing his own questions. His job is at stake, and his reputation, and most

37. George Boas, The Inquiring Mind (La Salle, 1959), p. 45.

FALLACIES OF QUESTION-FRAMING

25

important, his self-respect. If he substitutes a declarative for an inter-
rogative statement, then the result is literally a foregone conclusion. The
best will in the world won't suffice to keep him honest.

In historical writing, declarative questions tend also to be mimetic
questions, with a frequency which calls to mind an epigram (variously
attributed to Max Beerbohm and Herbert Asquith), that whether or not
history repeats itself, historians repeat each other. If historical research
were as empirical as it can be, then we might hope to see very large
heuristic hypotheses put to very small controlled tests. Now and then a
study appears which suggests that this dream is more than merely a mad
Utopian fancy. But many historians follow a different method. In common
practice, a general interpretation is fashioned by an essayist not as a
heuristic hypothesis but as an affirmative proposition. In the next twenty
years or so, a legion of gradgrinds manufacture monographs which
reify the essay, with a few inconsequential changes. Qualifications are in-
serted at the end of sentences, active verbs are changed to passive, pro-
nouns of indefinite reference are converted to proper nouns, and
footnotes are added at the bottom of the page. This process continues
until another essayist publishes another brilliant general interpretation,
and another generation of gradgrinds are wound up like mechanical rab-
bits and set to running about in ever-smaller circles. The result is a
dialogue between essayists and monographers which resembles the ex-
change between Hamlet and Polonius:

hamlet: "Do you see yonder cloud that's almost in the shape of a camel?"

polonius: "By the mass, and 'tis like a camel indeed."

hamlet: "Methinks it is like a weasel."

polonius: "It is backed like a weasel."

hamlet: "Or like a whale?"

polonius: "Very like a whale."

There is an example in the whales and weasels of Reconstruction
historiography. The first generation of academic interpretations of this
problem began with the publication of a set of interpretative Essays on
the Civil War and Reconstruction (1897) by Professor William Archi-
bald Dunning of Columbia University. Mr. Dunning and a few like-
minded gentlemen so dominated the American historical profession in
that era that a few disenchanted contemporaries spoke of his narrow
circle of acquaintances as the "history ring." 38

Under Dunning's direct control, or indirect influence, many state

38. J. G. de Roulhac Hamilton, ed., Truth in History (New York, 1937), p. xv. Hamil-
ton's introduction helps to explain the influence of Dunning, by demonstrating the
extraordinary deference structure which was sustained by the power of Dunning's per-
sonality, and the remarkable weight of his learning.

26

INQUIRY

studies of Reconstruction began to pour out of the seminar rooms of
Columbia and Johns Hopkins, most of which tended to reify the assump-
tions, interpretations, categories and prejudices of "the Old Chief." 38
These monographs were remarkably homogeneous in method and sub-
stance and extraordinarily consistent with Dunning's own opinions —
more so than if they were merely manifestations of a common Zeitgeist.
They tended to be elaborately detailed institutional studies, conceived as
political and constitutional history, informed by racist assumptions, and
colored by an antipathy to radicalism in all its forms and a particular
hostility to carpetbaggers, scalawags, and all that unsavory crew. They
were researched with care. But their authors willingly played Polonius
to Dunning's Hamlet.

A second wave of "revisionist" academic interpretations of Recon-
struction began with an article by a Negro historian, W. E. B. Du Bois,
"Reconstruction and Its Benefits," published in The American Historical
Review (15 [1910]: 781-99), and other work by Charles Beard. Though
Beard never wrote at length on Reconstruction, the chapters in his Rise
of American Civilization (2 vols., New York, 1927) were as influ-
ential as Dunning's work had previously been. The new generation of
monographers was a mixed bag of semi-Marxists, white liberals, and
Negro historians, whose interpretations tended to be more homogeneous
than their ethnic and ideological heterogeneity suggested. These "re-
visionists" were more interested in economic and social history than in
political and constitutional happenings and more apt to sympathize with
Negroes and carpetbaggers than with their conservative white opponents.
There was powerful inclination toward economic determinism in meta-
physics, to relativism in epistemology, and to egalitarianism in social
theory. 40

39. They included Hamilton Eckenrode, The Political History of Virginia During the
Reconstruction (1904); Walter L. Fleming, Civil War and Reconstruction in Alabama
(1905); James W. Garner, Reconstruction in Mississippi (1901); John R. Ficklen, His-
tory of Reconstruction in Louisiana (through 1868) (1910); Ella Lonn, Reconstruction
in Louisiana after 1868 (1918); J. G. de Roulhac Hamilton, Reconstruction in North
Carolina (1914); C. Mildred Thompson, Reconstruction in Georgia (1915); Charles W.
Ramsdell, Reconstruction in Texas (1910); Thomas S. Staples, Reconstruction in
Arkansas (1923); William W. Davis, The Civil War and Reconstruction in Florida
(1913); William Starr Myers, The Self-Reconstruction of Maryland (1909); John S.
Reynolds, Reconstruction in South Carolina (1905); James W. Fertig, Secessionism and
Reconstruction in Tennessee ( 1 898 ) ; James W. Patton, Unionism and Reconstruction in
Tennessee (1934); and J. P. Hollis, The Early Reconstruction Period in South Carolina
(1905).

40. Their works included Francis B. Simkins and Robert H. Woody, South Carolina
During Reconstruction (1932); Horace M. Bond, "Social and Economic Forces in
Alabama Reconstruction," Journal of Negro History 23 (1938): 290-348; Roger W.
Shugg, Origins of the Class Struggle in Louisiana (1939); Vernon L. Wharton, The Negro

FALLACIES OF QUESTION-FRAMING 27

A third wave of state studies is presently beginning to appear. The
most significant interpreters, the most influential teachers of graduate
students, are Kenneth M. Stampp at Berkeley, David Donald at Johns
Hopkins, Eric McKitrick at Columbia, and C. Vann Woodward at Yale,
all of whom have published many statements of their points of view.
Their approach is equally distinct from the politicoconstitutional history
of Dunning and the socioeconomic history of Beard. It consists princi-
pally of a sense of the complexity of an holistic configuration called
political culture, and a consciousness of the convolutions of human per-
sonality. Their histories tend to be intricate interactions of altruisms,
ideologies, "styles" of behavior, rational interests, irrational feelings,
and inchoate predispositions of various and sundry kinds. Their heroes
tend to be people who transcend institutional, intellectual, and psycho-
logical limits and who get things done. Their antiheroes (villains are
out of fashion) are blunderers and bunglers of all persuasions. A few state
studies have begun to appear: Joel Williamson, After Slavery: The
Negro in South Carolina During Reconstruction (1965); Joe M. Rich-
ardson, The Negro in the Reconstruction of Florida (1965); William
E. Parrish, Missouri Under Radical Rule (1965); Alan Conway, The
Reconstruction of Georgia ( 1 966) .

Any scholar who hopes for an empirical science of history cannot
be encouraged by this sorry tale of interpretative sequacity. But there
is also some rational reason for hope. Quantifiers will note that the
number and variety of master interpreters is increasing geometrically
in each generation, which suggests that freedom of scholarly inquiry
might be sustained by something like James Madison's notion of the
dynamics of republican freedom. Each generation of monographers is
a little more heterogeneous than the one before.

There is another hopeful tendency as well. Each generation does
not rewrite the history books; it revises them. From Dunning to Du Bois
to present practitioners, there is a process of refinement which is clearly
at work in Reconstruction historiography, a widening and deepening of
inquiry which transcends the reversals that relativism has taught us to
expect. Historians of Reconstruction have learned a little from their
predecessors — not much, but a little. And at the same time, they have
tacitly refined an ancient learning process,, which the Greeks were the
first to call heuristic.

This refining tendency suggests that an interpretative refinement

in Mississippi (1947); R. D. W. Conner, "The Rehabilitation of a Rural Common-
wealth," American Historical Review 36 (1930); 44-62; and Willie M. Caskey, Seces-
sion and Restoration of Louisiana (1938).

28 INQUIRY

might also be in order, in the terms of this essay. We began with a sim-
ple disjunction between declarative and interrogative statements. Maybe
we might more accurately conceive an intricate interactive relationship
between what is declarative and what is interrogative, within an opera-
tional question, which is always both declarative and interrogative in
some degree. Thus refined, the object becomes not the avoidance of
declarative hypotheses but careful control of the declarative part of
their structure, and at the same time, a careful control of the interrogative
aspect as well, so as to enlarge the empirical possibilities of historical
inquiry.

The fallacy of counterquestions is an attempt at a revision
which becomes merely a mindless inversion of an earlier interpretation
and a reiteration of its fundamental assumptions. It has been said that
there are two ways of manifesting an intellectual subservience to another
mind: slavish imitation and obsessive refutation. Both of these forms
of servility are regrettably common in historical scholarship. As revision-
ism grows more respectable, and even a prerequisite to a professional
career, an increasing number of historians are delivered into the latter
form of bondage.

Everyone has had some experience with a mind that "moves not
on wheels, but only on hinges," as Robert Hall complained of the un-
fortunate Dr. Chalmers. 41 A few seem seriously to believe that all minds
do and should work this way. A young radical American historian, Pro-
fessor Eugene Genovese, who is generally well disposed to dialectics,
appears to think that historians, like lawyers, ought to operate by an
adversary method. He endorses Santayana's eloquent statement that
"what kills spontaneous fictions, what recalls the impassioned fancy
from its improvisation, is the angry voice of some contrary fancy. Nature,
silently making fools of us all our lives, never would bring us to our
senses; but the maddest assertions of the mind may do so, when they
challenge one another. Criticism arises out of the conflict of dogmas." 42

This is surely false. A debate between two raving lunatics is un-
likely to issue in a triumph of reason. An argument between two patho-
logical liars is an improbable path to truth. An exchange between two
fools can scarcely be expected to end in a victory for wisdom. Adversary

41. William Gerard Hamilton, Parliamentary Logic, ed. Courtney S. Kenny (Cambridge,
1927), p. 82.

42. Quoted in Eugene D. Genovese, The Political Economy of Slavery (New York,
1965), p. 11.

FALLACIES OF QUESTION-FRAMING 29

methods may, perhaps, be appropriate to a courtroom, where the object
is the attainment of justice, but they are inappropriate to a seminar
room, where the purpose is the refinement of truth. A fight between
wild-eyed exponents of X and Y will help not at all if Z was in fact the
case, as it usually is. And between X and not- A" the difference is merely
a cipher, a nullity, a zero.

But there is something more specifically deficient about a counter-
question. If the original question, which is under attack, is mistaken,
then its basic assumptions are probably faulty. But a counterquestion,
in its reflexive inversion of the original, tends to repeat the original
assumptions, faults and all, and thereby to perpetuate the error. Counter-
questions repudiate conclusions but reiterate premises. The resultant
revision is objectionable not because it is revisionist but because its
revisionism is incomplete and superficial.

Consider, for example, the case of Charles Beard's An Economic
Interpretation of the Constitution of the United States (New York,
1913), and his most determined critic, Forrest McDonald. McDonald
has published two anti-Beardian books on the constitution: We the
People : The Economic Origins of the Constitution and E Pluribus Unum :
The Formation of the American Republic, 1776-1790 (Boston, 1965).
The first volume is an elaborate refutation of Beard's interpretation;
the second, a more positive statement of McDonald's views.

The labor McDonald performed was prodigious. Traveling some
70,000 miles in a battered, secondhand Plymouth, 43 he collected a
wealth of information in order to construct working models of the eco-
nomic and political structure of American society circa 1787 and to
compile brief economic biographies of the men who participated in the
Great Convention, and the state ratifying conventions as well.

But while McDonald's body was moving on wheels, his mind was
moving on hinges. Unlike the mind of many another revisionist, it did
so explicitly and deliberately in his first book: "The purpose of the present
work is to examine Beard's thesis as history. To do so, I have followed
a rather unorthodox method. For the purposes of this book I have
accepted, without qualification, Beard's system of interpretation and his
system of testing it." 44

That strategy was doubly dysfunctional to McDonald's inquiry.
First, it diverted his attention from the critical problem of logical and
epistemological and metaphysical deficiencies in the Beardian model.

43. We the People: The Economic Origins of the Constitution (Chicago, 1958), p. viii.

44. Ibid., p. vii.

30 INQUIRY

Second, when McDonald came to tell his own story of the Constitution,
the materials at hand were dictated by Beardian assumptions. The result,
in E Pluribus Unum, was that McDonald, in the words of one reviewer,
followed "a fairly Beardian line in interpreting the movement for the
Constitution. Except for a few special twists of his own, the general
outline follows the economic determinist pattern: failure of the impost
amendment of 1786, paper money, Rhode Island, Shays's Rebellion,
conservative reaction, etc." 45

Beard's book was built upon a conspiracy theory, and so was
McDonald's. The only difference was that McDonald's conspiracies
were more numerous and more complex. Beard began and finished with
a confused, clouded, contradictory, quasi-determinist economic deter-
minism, and so did McDonald. The only difference was that McDonald's
interest groups were more convoluted. For both authors, the dynamics
of the problem were analyzed in terms of "a handful of statesmen" who
were "impelled by an irresistible and illimitable compulsion to get
More." 46 The only difference is that McDonald recognized more states-
men and more ways of getting More than his predecessor had been able
to imagine.

There is at least one important way in which McDonald departs
from the Beardian model. He qualifies his account of the greed of the
founding fathers by giving some attention to gluttony and lust. Beard
delicately confined his investigation to pocketbooks; McDonald also
inspected the stomachs, glands, and genitals of the Founding Fathers.
But like Beard, McDonald tended to forget that men have minds and
hearts, and feet to stand on and spines to stand straight. In this respect,
McDonald became more Beardian than Beard.

Other parts of McDonald's second book, which are not merely
anti-Beardian but anti-Jensen and anti-Farrand in the same obsessive
way, suggest that his counterquestions were not merely methodological
but temperamental in their nature. His reviewer complained of a re-
flexive tendency toward the "categorical reversal of orthodox inter-
pretations." And though McDonald is an honest and an able scholar,
he tended also, in the process of turning things upside down, to give
the evidence a twist as well. A counterquestion tends to operate, in this
respect, as a declarative question. It does not open the inquiry, but
ends it.

There are many other historiographical examples of the fallacy of

45. E. James Fereuson in The William and Mary Quarterly, 3d ser. 23 (1966): 149.

46. E Pluribus Unum, p. 116.

FALLACIES OF QUESTION-FRAMING 31

counterquestions, and indeed, some very great and useful ones: Burck-
hardt and Huizinga, Tawney and Trevor-Roper, Weber and his many
critics, Hegel and Marx, English Whig historians and Namierites,
enemies and friends of the French Revolution, creators of the "Black
Legend" in Latin American historiography, and others who have re-
sponded with a whitewash. In American history, one thinks of Parring-
ton and Perry Miller on seventeenth-century New England, Thomas
Jefferson Wertenbaker and Wilcomb Washburn on seventeenth-century
Virginia, the Whig interpretation of the American Revolution and the
Imperial School, Henry Cabot Lodge and Edward Channing on the
politics of the early republic, Turner and Abernethy on the frontier,
Fiske and Jensen on the "critical period," Ulrich Phillips and Kenneth
Stampp on slavery, and others too numerous to mention.

In each of these historiographical pairs, the second man, or group,
is guilty of the fallacy of the counterquestion. In many cases, his work
was an improvement upon what went before, but it might have been
better still if counterquestions had been avoided and problems studied
without allowing assumptions to be established and problems to be
limited by earlier investigators.

The fallacy of tautological questions is the framing of ques-
tions in such a way that they are true by definition and cannot be
empirically contradicted without self-contradiction. A tautological ques-
tion is not really a question at all, but a declaration. Moreover, it is
doubly declarative, for it asks nothing and asserts the same thing twice.

There are three common varieties of tautological questions in
historical writing. The first and most common form of this error is a
simple proposition that all things which are P are in fact P. Calvin
Coolidge's famous first law of political economy is a familiar example.
"When people are out of work," he is alleged to have said, "unemploy-
ment results." But Coolidge's second law is not tautological, though
it may appear to be so. "The business of America is business," said he,
which is an equivocation. "Business" is used in two different senses to
sustain an argument that America's primary concern should be capital-
ism.

A historiographical example of this simple species of "P is P"
tautology is an article on radicalism and reform in The New York Times
Magazine, June 18, 1961, by Eric Goldman. Goldman asked, "Why do
some spectacular agitators forward their cause, and others do not?"
Before an audience of thousands, the author groaned and twisted upon

32 INQUIRY

his bed of conceptual agony, and at last, with a great rush of crimson
adjectives, he delivered himself of the following hypothesis: "All-out
agitators, to be successful, must be moving with history" (pp. 10-11).
But if the phrase "moving with history" means anything at all, it must
be something like a synonym for success. And if this is so, Mr. Gold-
man's profundity consists in a hypothetical proposition that "all-out
agitators, to be successful, must be successful."

Another example of the same sort of error occurs in The Rise of
Puritanism, a distinguished work of creative synthesis by William Haller.
But one of its major theses is a "P is P" proposition. "The cause of the
steady development of the centrifugal tendencies of Puritanism," writes
Mr. Haller, "was the Puritan fostering of individualism in religion in
the course of the accelerating democratization of English society." A
close critic of this work has complained that, given Haller's definitions,
he comes "near to saying that individualism was the cause of individual-
ism growing in conditions favorable to individualism." 47

A second kind of tautology consists in a proposition that all
things which are both P and Q are P. For example: "all red wagons are
red." Dr. Benjamin Spock somewhere sternly reminds American mothers
that "all babies are young." Dr. Barrington Moore solemnly reminds
American conservatives that all radical revolutionary change is violent —
which looks fine on first impression. But Barrington Moore's idea of a
radical revolution begins with violence as a central and limiting charac-
teristic. He singles out the most bloody chapter of American history,
the Civil War, as its most revolutionary chapter and ignores other
periods and processes which were more revolutionary and also more
radical in their effects, but less violent, and even nonviolent, in their
development. Moore's argument thus becomes an hypothesis that "all
violent radical revolutionary change is violent." 48 To forestall an accusa-
tion of ideological special pleading, I hasten to add a conservative ex-
ample. President William McKinley once declared that "our past has
gone into history" — this in a speech at Memphis, Tennessee, April 30,
1901, shortly before Mr. McKinley himself went into history.

The tautology "All things which are both P and Q are P" should,
however, be clearly distinguished from "All things which are P are P
and Q," which is tautological in "P is P" but more than merely a tautol-
ogy. Consider the statement attributed to Marshal Turenne, "He who has

47. William Haller, The Rise of Puritanism, 2d ed. (New York, 1957), p. 179; Richard
B. Schlatter, "The Problem of Causation in Some Recent Studies of the English Revolu-
tion," Journal of the History of Ideas 4 ( 1943) : 364 n.

48. Barrington Moore, Social Origins of Dictatorship and Democracy (Boston, 1966),
pp. 4, 20, 426-30, 505-8.

FALLACIES OF QUESTION-FRAMING 33

never made mistakes in war has never made war," which is not the
same as the statement "He who makes mistakes in war makes war" or
"He who makes mistakes in war makes mistakes."

A third general form of tautology is the assertion that something
is either P or it is not P. This sort of statement is very common in the
works of sixteenth-century historians, who were trained in dichotomous
reasoning. The writings of Machiavelli are chock-full of "P-or-not-P"
tautological hypotheses. "All states," he wrote in The Prince, ". . . are
either republics or monarchies." But he defined a republic in the usual
way as any state not monarchical. Similarly, Machiavelli hypothesized
often in the form, "The nobles are to be considered in two different
manners; that is, they are either to be ruled so as to make them entirely
dependent on your fortunes, or else not." 48

Consider also the following irreverent proposition:

First come I. My name is Jowett.
There's no knowledge but I know it.
I am Master of this College,
What I don't know isn't knowledge.

All three forms of tautology appear in hypotheses which are proposed
by Professor Bruce Mazlish in The Railroad and the Space Program
(Cambridge, 1 965 ) , pp. 34-35 : He begins by defining "social invention"
as "an invention which is technological (e.g., missile launching pads),
economic (e.g., involving large scale employment, widespread use of
materials), political (e.g., involving new forms of legislation and new
dispositions of political forces), sociological (e.g., affecting kinship
groups, communities, classes), intellectual (e.g., changing man's views
of space and time), and so forth" (p. 11). Then he proceeds to the
following hypotheses:

A. All social inventions are part and parcel of a complex — and have
complex results. Thus, they must be studied in multivariate fashion.

B. No social invention can have an overwhelming and uniquely deter-
mining economic impact . . . because no completely new innovation is possible
in reference to any set of economic objectives.

C. All social inventions will aid some areas and developments, but will
blight others.

D. All social inventions develop in stages, and have different effects
during different parts of their development.

E. AH social inventions take place in terms of a national "style," which
strongly affects both their emergence and their impact.

According to Mazlish's definition of "social inventions," all of his hypoth-

49. Niccold Machiavelli, The Prince and the Discourses, Modern Library ed. (New
York, n.d.), pp. 4, 37.

34 INQUIRY

eses are truistic and tautological — and all three forms of tautology are
included.

It has sometimes been argued that all hypotheses, or at least all
explanatory hypotheses, are tautological. 30 But this is surely a mistake.
Crane Brinton's explanatory generalization hypothesis that sociopolitical
revolutions tend to happen in societies which are relatively prosperous,
progressive, and "on the upgrade economically" is not a tautology.
Pieter Geyl's explanatory causal hypothesis that the difference between
the Flemings and Hollanders was caused not by differences of a racial
character but by the "geographical situation of Holland" is not a taut-
ology. Bernard Bailyn's explanatory descriptive motivational hypothesis
that Americans went to war against England partly because they feared
that a deliberate conspiracy was directed at the destruction of their free-
dom is not a tautology. Max Weber's explanatory paradigmatic hypothe-
sis that there was a functional interaction between the Protestant ethic
and the spirit of capitalism is not a tautology. An explanatory theoreti-
cal hypothesis such as the quantity theory of money, in the form of an
assertion that, all things being equal, the value of money is directly
proportional to the amount of money in circulation and the velocity
of its circulation — this is not a tautology. All of these explanatory hy-
potheses may be right or wrong, but they are not tautological. But to
insist, as some analytic philosophers of history would, that they are not
explanations is, itself, a tautology.

Tautologies can be exceedingly useful in the clarification of con-
ceptualization. If they are used intelligently, they can also be effective
rhetorical devices. But they ought not to be confused with heuristic,
open-ended questions of the sort which can be resolved by empirical
research.

The fallacy of contradictory questions is the framing of a ques-
tion which is false by definition and contradicts itself. If a truistic question
is a tautology, then a contradictory question is a naughtology.

One might pose a simple hypothetical example. An essayist asks
brightly, "What really happened in the summer of 1422, when, as
every schoolboy knows, an irresistible force met an immovable ob-
ject?" The question is contradictory, for if there are irresistible forces
then there can be no immovable objects. But our essayist is not dis-
mayed. With a flourish of hyphens and exclamation points he continues,
"I'll tell you what really happened in the horrible hot summer of 1422,

50. Eugene J. Meehan, Explanation in Social Sciences: A System Paradigm (Homewood,
111., 1968), p. 67.

FALLACIES OF QUESTION-FRAMING 35

when an irresistible force met an immovable object — there was one
Hell of a crash!"

This form of error is sometimes found in the bravura essays which
were much in fashion among historians of an older generation, who
worked themselves into impossible predicaments by framing contradic-
tory questions and then worked their way out by dodging them. The
object, I think, was to make historical writing appear difficult (and
thereby to magnify the apparent skill of the historian) by means of a
method which is impossible.

I am thinking particularly of the virtuoso pieces of A. J. P. Taylor,
the Paganini of historical prose, who likes to open an essay with a
paradox and to close it with a petitio, or else to begin with an insoluble
puzzle and end with an insidious quibble. One of his essays starts with
the question, "How has the continent of Europe escaped political uni-
fication? [Immovable object.] Everything in Europe seems to call for
it. [Irresistible force.]" Mr. Taylor proceeds to deal with this difficult
problem by rejecting both his major premises. With dark logic, but
with a brilliant display of rhetorical fireworks, he argues that the ir-
resistible force was not a force at all and that the immovable object
was actually in motion. And he caps off his exercise with a conclusion
which is altogether as perverse as his premises. Unification, Mr. Tay-
lor believes, did not come to "this most uniform of continents" because
"rejection of uniformity was the one thing uniform to the inhabitants
of Europe." 51

Mr. Taylor does it again in a famous essay on Napoleon III. He
begins with a conventional complaint that the more we learn about
Napoleon III, the less we really know. "The more we strip off [the]
disguises, the more new disguises appear. Such was Louis Napoleon,
the man of mystery. Conspirator and statesman; dreamer and realist;
despot and democrat; maker of wars and man of peace; creator and
muddler; you can go on indefinitely. . . ."

But Taylor is not discouraged. He piles one puzzle upon another
in a great towering pyramid of cleverness and confusion. The reader
scarcely has time to study any one of them before it disappears beneath
others. In the end, there is a conundrum: "It was easy to be against
Napoleon when he turned out to be the man of Sedan. It was his doom
that he was branded from the start, and branded in history, as the
man of December," which is ambiguous, obscure, and irrelevant to
the question which Taylor originally raised. 52

51. A. J. P. Taylor, "Napoleon and Gentz," in From Napoleon to Lenin: Historical
Essays (New York, 1966), pp. 12-20.

52. "The Man of December," ibid., pp. 76-81.

36 INQUIRY

It is important, however, to distinguish clearly between a contra-
dictory question and a paradoxical question — which is to say, between
a contradiction and a seeming contradiction. The latter is much ex-
ploited by historians, not merely for rhetorical purposes but for sound
research purposes as well. The works of an American historian, David
W. Noble, are a case in point: The Paradox of Progressive Thought
(Minneapolis, 1958) and Historians Against History (Minneapolis,
1965). The first title could read, "Progressives against Progress"; the
second, "The Paradox of Historical Thought." Noble's answers may be
mistaken, but his questions are workable.

The fallacy of "potentially verifiable" questions consists in
the mistaken and mischievous idea that a division of labor is both
possible and desirable between historians who identify hypotheses which
might be put to the test and other historians who test them. The mis-
take of this method lies in an attempt to separate two interdependent
parts of a single process. Any such attempt to divide history, like
physics, into "theoretical" and "experimental" branches would probably
impoverish both aspects of a historian's task.

Question-framing cannot be undertaken independently of question-
answering, for no hypothesis can be demonstrated to be potentially veri-
fiable except in the degree to which it has been partially verified. When
a historian says, "I have reason to believe that the question 'Was X
or Y the case?' can be answered," he means that he has some evidence
to suggest that either X or Y was the case, but that the evidence is in-
complete or inconclusive. Moreover, so vastly complex is the process
of verification, and so utterly unpredictable are the obstacles which lie
hidden along the way, and so intimate is the functional relationship
between the design of questions and the attempt to resolve (and refine)
them, that the two processes cannot be separated, except at a heavy
cost to the quality of conceptualization and research which is accom-
plished.

This truth does not need to be taught to working historians, who
have learned it by bitter experience. But it is, perhaps, a lesson which
they can teach to a distinguished sister discipline. In 1949, a prominent
sociologist wrote of his own colleagues that "sociologists (including
this writer) may discuss the logical criteria, of sociological laws without
citing a single instance which fully satisfied these criteria." 53

Professor Merton's statement applies with greater accuracy to

53. R. K. Merton, Social Theory and Social Structure (Glencoe, 111., 1949), p. 92.

FALLACIES OF QUESTION-FRAMING 37

sociology today than it did twenty years ago. His discipline has become,
if anything, even less empirical than it used to be. "The weakness of
much social thought, it seems to me," writes an English historian, "is
that it is so largely concerned with packing its bag (or even with work-
ing out a general theory about the way in which a bag should be
packed) for a journey which is never taken." 54

To complain that sociologists have habitually tended to commit
the fallacy of potentially verifiable questions, is not, however, to issue
an invitation to another interdisciplinary slugging match, from which
nothing good can come. E. H. Carr's fine epigram can bear repeating
many times: "The more sociological history becomes, and the more
historical sociology becomes, the better for both. Let the frontier
between them be kept wide open for two-way traffic." 53 But let us
hope that the two-way traffic will keep to the right side of the road.
If sociological history and historical sociology are conceived as a
combination of the conceptual sophistication of the best sociologists
and the dogged if often undirected empiricism of the best historians,
then the prospects are very bright indeed. But one might also imagine
an interdisciplinary effort which combined the worst of both worlds —
the stupidity of historians and the ignorance of sociologists.

This unfortunate tendency is apparent in the early work of Pro-
fessor Lee Benson, a historian who has tended to borrow the vices of
other disciplines and to surrender the virtues of his own. Prominent
among his takings from sociology was the idea of the "potentially
verifiable" question, which he elevated into an explicit method and
urged upon his colleagues as a new tool of historical inquiry. 56 Happily,
this well-intended but ill-founded advice has not been needed, even by
historians who have adopted some of Professor Benson's more con-
structive suggestions. We shall have occasion, in a later chapter, to
consider the toll which this erroneous method has taken of his own
work.

The potentially verifiable question is perhaps a manifestation of
Abraham Kaplan's myth of methodology. 57 Let us hope that historians
will hear the advice of a sociologist, Professor T. H. Marshall, who has
written. "There was a compelling persuasiveness about the famous cry
— 'Give us the tools and we will finish the job.' One may be forgiven

54. Alfred Cobban, The Social Interpretation of the French Revolution (Cambridge,
1964), p. 23.

55. E. H. Carr, What Is History?, p. 84

56. Lee Benson, "Research Problems in American Political Historiography," in Mirra
Komarovsky, ed.. Common Frontiers of the Social Sciences (Glencoe, III., 1957),
pp. 113-81.

57. See above, p. xx.

38

INQUIRY

for responding less eagerly to the scholar, be he sociologist or anything
else, who says — 'Give me a job, and I will spend the rest of my life
polishing the tools.' " 58

These eleven common fallacies of factual question-framing
suggest a few affirmative axioms, which, however simple they may be,
are often honored in the breach.

First, a proper historical question must be operational — which is
merely to say that it must be resolvable in empirical terms. This simple
requirement has been elevated into an ism by philosophers who speak
of "operationism," or "operationalism," and define it as "the demand
that the concepts or terms used in the description of experience be
framed in terms of operations which can be unequivocally performed." 59
But this is merely a common-sense notion.

Second, a question should be open-ended, but not wide-open. It
should dictate the kinds of facts which will serve to solve a problem,
without dictating the solution itself. It must be a genuinely interrogative
statement, but at the same time it must guide the inquiry through masses
of information. If it does not perform the latter function, the historian
will share Alice's confusion, as she went a-wandering in Wonderland:

"Cheshire-Puss," she began, rather timidly. . . . "Would you tell me,
please, which way I ought to go from here?"

"That depends a good dsal on where you want to get to," said the Cat.

"I don't much care where — " said Alice.

"Then it doesn't matter which way you go," said the Cat. 80

One way to balance the difficult dual requirement of freedom and
control is to begin with a cluster of questions, and for each of them, a
cluster of answers which are generated by hunches and preliminary
explorations and refined into alternative hypotheses, which can be en-
larged or altered as research may require. These clusters of questions
and hypotheses can and indeed must be designed in such a way as to
neutralize a predisposition to actualize any one of them. There is no
other way to keep honest and at the same time to keep one's momentum
through masses of source material:

Third, a question should be flexible. A historian must learn to
resist that form of debilitation which has been called "hardening of the

58. T. H. Marshall, Sociology at the Crossroads (London, 1947), p .19.

59. Carl G. Hempel, Fundamentals of Concept Function in Empirical Science (Chicago,
1952), pp. 39-50; Percy Bridgman, The Logic of Modern Physics (New York, 1948),
p. 5, passim.

60. The Complete Works of Lewis Carroll, p. 72.

FALLACIES OF QUESTION-FRAMING 39

categories." 61 He must learn to conceive his questions and hypotheses
as approximations, which are open to infinite refinement. A question
cannot be framed wholly from historical sources — a historian must
start with something. But it can be adjusted and amended, revised and.
ramified.

Fourth, a question must be analytical, which is to say that it must
help a historian to break down his problem into its constituent parts,
so that he can deal with them one at a time. It has been wisely and
wittily observed that "the only practical problem is what to do next." 42
A proper question must serve to assist in this process, by separating
sequential steps of inquiry.

Fifth, a question must be both explicit and precise. Its assumptions
and implications must be spelled out in full detail, not merely for the
sake of the reader, but for the sake of the researcher himself. Nothing
is more deleterious and more absurd than the common tendency of some
historians to confuse open-mindedness with imprecision, and flexibility
with bef uddlement, and wisdom with obscurity.

Finally, a question must be tested. No hypothesis can be conceived
as "empirically verifiable" except in the degree to which it is verified.
"Questions are not put by one man to another man, in the hope that the
second man will enlighten the first man's ignorance by answering them,"
wrote R. G. Collingwood. "They are put, like all scientific questions,
by the scientist himself." 63

61. F. H. Underhill, quoted in Toynbee, Reconsiderations (New York, 1961), p. 1.

62. I. J. Good, The Scientist Speculates: An Anthology of Partly Baked Ideas (New
York, 1962), p. 213.

63. Collingwood, The Idea of History, p. 274; and Autobiography, A30-31, 37.

CHAPTER II

FALLACIES OF FACTUAL
VERIFICATION

One of the first duties of man is not
to be duped.

— Carl Becker

The best historical questions are no better than the answers which
they generate — true answers, which are the object of all empirical in-
quiry. It is no easy matter to tell the truth, pure and simple, about past
events; for historical truths are never pure, and rarely simple. And
the process of historical truth-telling itself is even more intricate than
the truths which historians tell. Every true statement must be thrice
true. It must be true to its evidence, true to itself, and true to other his-
torical truths with which it is colligated. Moreover, a historian must not
merely tell truths, but demonstrate their truthfulness as well. He is
judged not simply by his veracity, but by his skill at verification.

Anyone who has followed the progress of historical scholarship in
the past generation will probably agree that this fundamental skill of
factual verification has not been sufficiently attended to. The incidence of
simple factual error in academic history today is nothing short of ap-
palling. Two conventional attitudes are, I think, commonly responsible.
First, the requirement is so fundamental that it is too often taken for
granted. Few graduate programs in history deliberately teach students
how to discover particular truths, and how to demonstrate truthfulness.
It is merely assumed that historians will do so. An English scholar
writes condescendingly,

It is no doubt important to know that the great battle was fought in 1066
and not in 1065 or 1067, and that it was fought at Hastings and not East-
bourne or Brighton. The historian must not get these things wrong. But
when points of this kind are raised, I am reminded of Housman's remark
that "accuracy is a duty, not a virtue." To praise a historian for his accuracy

FALLACIES OF FACTUAL VERIFICATION

41

is like praising an architect for using well-seasoned timber or properly
mixed concrete in his building. It is a necessary condition of his work, but
not his essential function. It is precisely for matters of this kind that the
historian is entitled to rely on what have been called the "auxiliary
sciences" of history — archaeology, epigraphy, numismatics, chronology, and
so forth. 1

This haughty attitude is a very unfortunate habit, which many
eminent historians have fallen into. The apparent fact that historians
often assume that particular statements will be accurate, or that some-
body else is responsible for their accuracy, may serve to explain why
so many historical statements are in fact inaccurate. There are no
auxiliary sciences to which a historian can conveniently dispatch his
verification problems, and none to relieve him of his responsibility in
this respect.

Carr's argument is also objectionable because it simplifies a vastly
complicated process, which is inseparable from historical inquiry and
functionally related to its most intricate conceptual parts. An inspec-
tion of the following fallacies will, I hope, suffice to demonstrate the
fact that erroneous understandings of the verification process have
caused able historians to adopt dysfunctional methods and procedures,
which are utterly destructive of empirical scholarship.

Second, historical accuracy has also been diminished, during the
past generation, by the progress of historical relativism. This absurd
and pernicious doctrine became a popular delusion in the 1930s, when
many historians suddenly discovered the disturbing fact that history
was something which happened to them. The effect of that uncomfortable
revelation was reinforced by the failure of history to develop as satis-
factorily as the natural sciences, and by the fact that a good deal of
humbug had been proclaimed to the world as the objective truth of
history, and by a great falling away from first principles in the Western
world, and especially in the Western democracies.

Relativism became particularly powerful in the United States,
where the polemics of Charles Beard and Carl Becker reached many
historians born in the generation 1890-1930, and indirectly, much of the
general public as well. The way was prepared by a profound Pyrrhonism
in the popular culture which was perfectly exemplified by Poor Richard's
saying that "historians relate, not so much what is done, as what they
would have believed," or in General George Meade's Philistine assertion
that "I don't believe the truth will ever be known, and I have a great
contempt for history."

But historical relativism was also an international phenomenon,

1. E. H. Carr, What Is History? (New York, 1962), p. 8.

42 INQUIRY

which has appeared in formal treatises by a great Chilean historian, a dis-
tinguished French scholar, and an Hungarian-born historical sociologist,
to name but a few. 2 A full-fledged cultural history of this intellectual
movement would embrace a major school of German historical and
philosophical thought, the epistemology of Fascism, 3 Stalin's hostility
to "archive rats" and "bourgeois objectivism," the great Japanese movie
Rashomon and several popular anthologies of Hindu folk legends, the
novels of Aldous Huxley, Thomas Mann and Andre Gide, the essays
of Renan and Croce, the poetry of Edward Arlington Robinson and E.
E. Cummings, the theology of Reinhold Niebuhr, and the aesthetics
of abstract art.

In poetry and painting, relativism was a harmless error, but when-
ever it appeared in historical scholarship, the results were disastrous. The
fallacies committed by relativists have been sufficiently exposed by
philosophers. 4 They require no laborious discussion here. There is, of

2. Francisco A. Encina, La Literatura Histdrica Chilena y el Concepto Actual de la
Historia (Santiago, 1936); Raymond Aron, Introduction a la Philosophie de I'Histoire
(Paris, 1938); Karl Mannheim, Ideology and Utopia: An Introduction to the Sociology
of Knowledge (New York, 1936).

3. Svend Ranulf, Hitlers Kampf gegen die Objektivitat (Copenhagen, 1946).

4. A good brief discussion of Beard's relativism is in Arthur C. Danto, Analytical
Philosophy of History (Cambridge, 1965), chap. 6. Karl Mannheim's "relationism,"
which was actually a form of relativism even less defensible than Beard's, is refuted in
Charles Frankel, The Case for Modern Man, 2d ed. (Boston, 1959), chap. 7. Every his-
torian can profit by a reading of these two critiques. In brief, the errors of relativism con-
sist in the following:

First, there is a confusion between the way knowledge is acquired and the validity
of that knowledge. An American historian may chauvinistically assert that the United
States declared its independence from England in 1776. That statement is true, no
matter what the motives of its maker may have been. On the other hand, an English
historian may patriotically insist that England declared its independence from the United
States in 1776. That assertion is false, and always will be.

Second, relativism mistakenly argues that because all historical accounts must be
partial in the sense of incomplete, that they must also be partial in the sense of false. An
incomplete account can be an objectively true account; it cannot be the whole truth.
In this respect, the relativists continued to bootleg the idea of telling the whole truth
in their work.

Third, relativism makes false distinctions between history and the natural sciences.
Beard in particular did this, and his error consisted in rendering a special judgment upon
historical science for its use of hypotheses, etc., which are also characteristic of natural
science. Arthur Danto comments, "It is as though a man were to lament that it is a sad
thing to be a Frenchman, for all Frenchmen die. . . . History is no more and no less
subject to relativistic factors than science is" (Danto, p. 110). Mannheim made the
same error in Ideology and Utopia (p. 79).

Fourth, relativists all argued that they and their friends were exempt from relativism
in some degree. Thus, Beard's special pleading for an economic interpretation; and
Mannheim's, for the intelligentsia. Both scholars were inconsistent, and understandably
so. Cushing Strout has observed that "a consistent relativism is a form of intellectual

FALLACIES OF FACTUAL VERIFICATION

43

course, something that is profoundly right in relativism. It is true that
history is something which happens to historians. And it is correct to
argue that no historian can hope to know the totality of history as it
actually happened. But it is wrong to conclude that objective historical
knowledge is therefore impossible. The conventional wisdom of con-
temporary historiography still consists in the common idea that "a
historian cannot know what really happened, but he has a duty to try."
It is not surprising that historians have not tried very hard. Among the
results are the following fallacies.

The fallacy of the pseudo proof is committed in a verifica-
tion statement which seems at first sight to be a precise and specific
representation of reality but which proves, on close inspection, to be
literally meaningless.

Consider the work of a single scholar — Carl Bridenbaugh of Brown
University, an able and distinguished author of many monographs about
American colonial history, a president of the American Historical
Association, a pioneer in the development of American social history,
and one of the best and most useful men in his historiographical genera-
tion. But the methodological faults of that generation frequently appear
in his work, which is grossly impressionistic in style and substance. Many
of his conclusions are sustained by one fact at most, and some by a
pseudo fact.

In a book entitled Cities in Revolt (New York, 1955), Bridenbaugh
argued that Bostonians were heavily taxed in the period 1743-1760.
His evidence consisted in an exclamatory assertion that "at the close of
this period the levy on the 'Estates Real and Personal' of Bostonians
amounted to 1 3s. 6d. in the pound, or 67 percent!" 5 But this statement,
in itself, tells the reader nothing. Were those thirteen shillings and six-
pence extracted from a pound of property at market value or from an
assessed valuation of estates? Bridenbaugh doesn't tell us. Let us as-
sume the latter, which was probably the case. If so, what were the
assessment rates in proportion to real value — 100 percent? 50 percent?

suicide" (The Pragmatic Revolt in American History: Carl Becker and Charles Beard
[New Haven, 1958], p. 84).

Finally, the idea of subjectivity which the relativists used was literal nonsense. "Sub-
jective" is a correlative term which cannot be meaningful unless its opposite is also
meaningful. To say that all knowledge is subjective is like saying that all things are
short. Nothing can be short, unless something is tall. So, also, no knowledge can be
subjective unless some knowledge is objective. (See Christopher Blake, "Can History
be Objective?" Mind 72 [1955]: 61-78.)
5. P. 7.

44 INQUIRY

5 percent? If rates were high, then Bostonians were very heavily taxed,
in whatever year Bridenbaugh found his figure. But if they were low,
then the Boston tax rate might have been absurdly small. Bridenbaugh's
"fact" helps not at all to clarify the confusion. As it is presented to the
reader, it has no more evidential value than the exclamation point
which ends his sentence.

This may serve as a specimen of one sort of pseudo fact, in
which a relative quantity is stated in absolute terms, without a clarifi-
cation of its reference. Another example of precisely the same species
appears in Bridenbaugh's The Colonial Craftsman (2d ed. [Chicago,
1961]), which is studded with impressionistic statements of prices,
wages, profits, and investments, all expressed in pounds and shillings.
But pounds and shillings in what currency? Bridenbaugh helpfully
identifies one figure as pounds sterling, and another as pounds, Massa-
chusetts currency, which had a different and changing value. The rest
of his figures are for the most part unspecified, and therefore unclear.
What were these pounds worth, in purchasing power? Bridenbaugh
cheerfully provides two conflicting estimates, one of which equates a
"pound" to $30; the other, to $25, at "present day values." But which
"pound" and what "present day" — the date of the first edition (1950)
or the second (1961)? Every point pounded is confusion compounded.
Bridenbaugh's method adds something to the weight of the book but
nothing to its value."

A second species of pseudo fact is the reversible reference — a
chameleonlike statement which changes its color with its context and
which might variously be used to prove the proposition that X is the
case or that not-A' is the case, as the author wishes. Here again, Briden-
baugh's books provide a plenitude of examples. In Cities in the Wilder-
ness ([New York, 1955], p. 18), he asserts that the streets of colonial
towns were strewn with trash: "Casting rubbish and refuse of all kinds
into the streets without let or hindrance was a confirmed habit of both
English and American town-dwellers," he wrote confidently. It may have
been so, but his supporting evidence for Manhattan consists of three
reversible references — that a law was passed against littering in New
Amsterdam in 1657, that the law was enforced upon an early American
litterbug named John Sharp in 1671, and that provision was made in
1670 for weekly trash removal by the car men of the city. Each of these
impressionistic snippets of pseudo-factual information is consistent with
a thesis that ( 1 ) the streets of New Amsterdam were knee-deep in trash;
or (2) the streets of New Amsterdam were kept spotlessly clean by the

6. Pp. 42, 44, 128, passim.

FALLACIES OF FACTUAL VERIFICATION

45

tidy Dutch inhabitants, by means of laws which were enforced and by
regular trash removal; or (3) any statement between these two extremes.
The problem of the reversible reference rises whenever a historian tries
to draw an inference directly from law to life. A law against X can be
interpreted as evidence for the existence of X or for its nonexistence.

Another error of this same sort appears in Bridenbaugh's Myths
and Realities (Baton Rouge, La., 1952), where he attempts to demon-
strate that eighteenth-century Virginians were not much interested in
music. To sustain this statement, he quotes Thomas Jefferson, who
complained that he lived in a country where music "is in a state of
deplorable barbarism." 7 But Jefferson was part of this society, and his
statement is primary evidence that at least one eighteenth-century Vir-
ginian was interested in music, and secondary evidence of the barbarism
of his neighbors. It might be a more accurate indicator of the quality of
Jeffersonian expectations than of the musicological decadence of the
Old Dominion. In the absence of clarifying evidence it is meaningless.
Whenever a historian quotes an allegation by a member of group A, to
the effect that A is not sufficiently interested in B, he is in danger of
perpetrating a pseudo fact.

A third kind of pseudo fact is close to another fallacy in the follow-
ing chapter. It consists in the precise quantification of imprecise entities.
Bridenbaugh solemnly tells us that "at this time [no date] at least eight
drawing schools drew their support from the wish of Charles Town's
young ladies to learn the rudiments of a fashionable art." 8 This "fact" is
at first sight an impressive testimony to the painterly propensities of
Charleston belles — until one begins to wonder what precisely a drawing
school could be. Bridenbaugh's statement could refer to eight splendid
institutes with white marble steps and a faculty of French masters. Or it
could mean eight old women with a community paint brush. Surely the
truth is somewhere in the middle. But Bridenbaugh leaves us merely in a
muddle.

The fallacy of the irrelevant proof consists in asking one
question and answering another. Suppose, for example, that a historian
asks, "Was Senator X a thief?" And suppose, moreover, that Senator X
was a very great thief. But our hypothetical historian, who is an admirer
of Senator X, proceeds to prove that Senator X had often declared that
honesty was the best policy. He demonstrates that Senator X was acquit-

7. P. 48.

8. Myths and Realities, p. 111.

46 INQUIRY

ted of theft by a jury (without mentioning that all the jurymen became
postmasters immediately after the trial). He publishes an affidavit in
which Senator X's mother solemnly swore that her son could never be
a thief. He establishes that Senator Y was a bigger thief than Senator X,
and that the Senate itself was a den of thieves. He shows that Senator
X only stole from the government, and was kind to his children, and
faithful to his wife, and loyal to his party. He proves that Senator X used
some of his money to pay for an operation which was desperately needed
by a crippled orphan in Cincinnati, Ohio. He argues, in a learned Keynes-
ian disquisition, that Senator X was a big spender in a state where
money was scarce, and that his spending, compounded by many mul-
tipliers and linkage effects, brought prosperity to thousands, and fac-
tories, jobs, schools, and churches. All of these statements may in fact
be true, and yet Senator X remains a thief.

Something very similar to this hypothetical example actually ap-
peared in the American Historical Review over the signature of a great
and gifted historian — Allan Nevins. The point at issue was the validity
of Henry Demarest Lloyd's allegations against John D. Rockefeller and
the Standard Oil Company, in a polemic called Wealth Against Com-
monwealth. Allan Nevins had, at one time, shared Lloyd's point of view.
In his biography of Grover Cleveland, Nevins characterized Wealth
Against Commonwealth as "a searching exposure, amply buttressed by
detail," and an accurate rendering of a "sordid record of business piracy,"
in "more than five hundred calm, unemotional pages." But a few years
later, Nevins published a biography of Rockefeller in which he con-
demned Lloyd's work as "almost utterly worthless" and even dishonest.

One of Nevins's colleagues, Chester McA. Destler, replied with a
critical essay in the American Historical Review, in which he defended
Lloyd's book against Nevins's strictures. Specifically, Destler checked
some 420 footnotes in Wealth Against Commonwealth, and found all
but ten of them to be fair and accurate (the ten exceptions were inaccu-
rate in insignificant ways). He checked 241 unsupported statements by
Lloyd, and found 229 of them to be true and just. Destler concluded,
persuasively, that Lloyd was correct in charging Standard Oil and Rocke-
feller with the bribery of public officials, the corruption of judges and
juries, the negotiation of illegal pooling agreements, shoddy treatment of
inventors, and sundry acts of espionage, coercion, fraud, and physical vio-
lence against competitors. He conceded that Lloyd had wrongly con-
demned Rockefeller for having allegedly bilked a Cleveland widow out of
her life savings. But otherwise, he insisted, Lloyd's indictment of the
methods of Standard Oil and Rockefeller was substantially accurate and
just.

FALLACIES OF FACTUAL VERIFICATION

47

To this, Nevins made the rebuttal that John D. Rockefeller was a
"hard-working, home-loving, religious" man, that he founded the Uni-
versity of Chicago, and that "his house was filled with missionaries and
social workers." Nevins argued at some length that Rockefeller did not
defraud the Cleveland widow out of her property (a point which Destler
had granted). To a charge that Standard Oil bought a seat in the U.S.
Senate for Henry B. Payne by bribing the Ohio legislature, Nevins seri-
ously replied that the Ohio legislature made an investigation of this
charge and found it to be unsustained. To the accusation that the Stand-
ard Oil Company had engaged in physical violence against its rivals and
bribed a jury to acquit the perpetrators or to let them off with a trivial
fine, Nevins replied that the alleged perpetrators had been acquitted, or
let off with a trivial fine. "Men guilty of trying to blow up a factory are
not treated so lightly," Nevins piously declared. 9

$*> The fallacy of the negative proof is an attempt to sustain a
factual proposition merely by negative evidence. It occurs whenever a
historian declares that "there is no evidence that X is the case," and then
proceeds to affirm or assume that aot-X is the case. He may have spent
all the years of his youth in the Antiquarian Society, feverishly seeking
the holy X and never finding it. He may have examined every relevant
scrap of evidence in every remote repository, without reward. He and
every other reasoning being on this planet may know in their bones that
not-AT is the case. But a simple statement that "there is no evidence of
X" means precisely what it says — no evidence. The only correct empirical
procedure is to find affirmative evidence of not-X — which is often dif-
ficult, but never in my experience impossible.

An example of the fallacy of negative proof occurs in a recent
essay by Mary and Oscar Handlin, The Popular Sources of Political
Authority: Documents on the Massachusetts Constitution of 1780 (Cam-
bridge, Mass., 1966). Constitution-making in Massachusetts was a
complex process, of central relevance to any interpretation of political
development in the early American republic. Great clouds of controversy
swirl around this chapter of our history, and they have grown dense and
dark with time. Among many urgent and unresolved problems is the
question of social and cultural correlates with political commitments.

The Handlins, however, airily dismissed this issue with a simple
statement that "the editors have been unable to establish a meaningful
correlation between the social conditions of specific towns and the at-

9. Chester McA. Destler, "Wealth Against Commonwealth, 1894 and 1944," Ameri-
can Historical Review 50 ( 1944-45 ) : 49-72, and reply, pp. 679 ff.

48 INQUIRY

titudes expressed in the responses [to constitutional questions]." They in-
sisted that "no conclusion was possible from an examination of the
returns." Evidence collected by other historians has suggested that there
was at least a clear pattern of denominational response to those parts of
the Constitution which concerned religion — towns with large Baptist
populations being averse, unsurprisingly, to a state-supported Congrega-
tionalist establishment. But Mr. and Mrs. Handlin simply state that
"even in this matter the correlation was not exact," and drop the ques-
tion. 10

The authors, in short, declared categorically that there is no
evidence that X is the case, instead of presenting affirmative evidence
of not-A r . But affirmative evidence of not- A' could have been obtained
(if in fact not- A' was the case) by the calculation of correlation coeffi-
cients between, for example, the number of Baptist churches in the towns
and the town votes on religious provisions of the Constitution. A low
correlation would stand as positive evidence of not- A'. Something similar
might have been done with respect to the wealth of the towns, their
demographic characteristics, and their political identities.

Had the Handlins collected such positive evidence, in my opinion,
their thesis would have collapsed. It is exceedingly likely that in fact X,
rather than nol-X, was the case, and that many correlations between
social, religious, and economic variables on the one hand, and political
variables on the other, can be found. But the Handlins, able and honesf
historians, seem not much interested in finding correlations of this sort,
perhaps partly because such findings would contradict some of their most
cherished beliefs about American history in particular, and humanity in
general.

A good many scholars would prefer not to know that some things
exist. But not knowing that a thing exists is different from knowing that
it does not exist. The former is never sound proof of the latter. Not
knowing that something exists is simply not knowing. One thinks of Alice
and the White Knight:

"I see nobody on the road," said Alice.

"I only wish / had such eyes," the King remarked in a fretful tone.
'To able to see Nobody! And at that distance too!" 11

The fallacy of the presumptive proof consists in advancing a
proposition and shifting the burden of proof or disproof to others.

10. Pp. 933, 475.

1 1. The Complete Works of Lewis Carroll, Modern Library ed. (New York, n.d.), p. 223.

FALLACIES OF FACTUAL VERIFICATION

49

An example comes from Erik H. Erikson's fine monograph Young
Man Luther. That controversial psychoanalytic interpretation of the re-
former is grounded in an assumption that many specific Freudian and
post-Freudian insights into human behavior are equally valid in Erik-
son's era and in Luther's. The author writes, with specific reference to
the Oedipus complex, "I will not discuss here the cultural relativity of
Freud's observations nor the dated origin of his term; but I assume that
those who do wish to quibble about all this will feel the obligation to
advance specific propositions about family, childhood and society which
come closer to the core, rather than go back to the periphery of the riddle
which Freud was the first to penetrate." 12

Aside from Erikson's unfortunate condemnation of his critics in
advance, as mere "quibblers," this passage, which is centrally important
to the validity of his interpretation, is seriously objectionable. There is
a burden of responsibility which rests squarely upon Erikson, and not
upon his quibbling critics, to advance specific propositions which come
closer to the core.

The raw application of timeless ideas about identity crises and
Oedipus complexes — which are surely dependent in some degree upon
family structure, child-rearing practices, and sundry other changing cul-
tural conditions — to a man whose cultural environment was far dif-
ferent from Freud's, or Erikson's, without a satisfactory attempt to allow
for those differences, is a disconcerting quality of Erikson's book.

£»> The fallacy of the circular proof is a species of a question-
begging, which consists in assuming what is to be proved. A hypothetical
example might help to clarify the point. A researcher asks, "Do gentle-
men prefer blondes?" He discovers that Smith, Jones, and James prefer
blondes, and tacitly assumes that Smith, Jones, and James are therefore
gentlemen. He concludes that three gentlemen out of three prefer blondes,
and that the question is empirically established, with a perfect correla-
tion. His argument runs through the following stages:

Inquiry: Do gentlemen prefer blondes?

Research: Smith, Jones, and James prefer blondes.

(Tacit Assumption) : Smith, Jones, and James are gentlemen.

Conclusion: Therefore, gentlemen prefer blondes.

Absurd as this fallacy may appear in a hypothetical way, it is ex-
ceedingly common in empirical scholarship. Consider the following case

12. Erik H. Erikson, Young Man Luther: A Study in Psychoanalysis and History, 2d ed.
(New York, 1962), p. 257.

50 INQUIRY

by a historian of old Ireland. "The question has been posed," writes
Jeremiah O'Sullivan, "whether St. Patrick and Palladius were one and
the same person. Tirechan in his Memoir of St. Patrick states that Pal-
ladius also had the name of Patrick and changed his name as Patrick
did. The best proof that they were not one and the same seems to be that
St. Patrick died in Ireland in 461, whereas the time and circumstances of
Palladius's death are uncertain." 13

This may indeed be the best available proof, but it is not good
enough to carry the question. If Palladius were known to have died in
470, somewhere in Germany, then we would have very good evidence
that he and Patrick were not the same. But the uncertain circumstances
of Palladius's death are clarified in a way which is consistent with
O'Sullivan's hypothesis only if we assume that his hypothesis is correct.

Other examples appear in a provocative new book on English
Puritanism by Michael Walzer, who seeks to demonstrate by an appeal
to historical evidence that a certain complex of ideas and feelings specifi-
cally attached to Puritanism. But when he finds an expression of these
ideas and feelings, he assumes that the thought of the man who expressed
them was "Puritanical" simply because he expressed them. This hap-
pens even in the case of men who are known to have been hostile to the
movement conventionally called Puritanism. 14

In other words, Walzer argues that Puritans were men who thought
X, but he attempts to prove his thesis by assuming that men who thought
X were Puritan. This is a convenient way to collect vast quantities of
evidence. But it might make the Pope himself into a Puritan. In the same
fashion, Walzer seeks to prove that certain other patterns of thought were
abandoned by the Puritans. But when Puritans expressed those thoughts,
Walzer explains those expressions away by arguing that Puritans really
didn't mean them. 16

The fallacy of the circular proof, or question-begging, also has its
converse, which is described by Augustus de Morgan.

There is an opponent fallacy to the petitio principii, which, I suspect, is of
the more frequent occurrence [he writes]. It is the habit of many to treat
an advanced proposition as a begging of the question the moment they
see that, if established, it would establish the question. Before the advancer
has more than stated his thesis, and before he has time to add that he
proposes to prove it, he is treated as a sophist on his opponent's perception
of the relevancy (if proved) of his first step. Are there not persons who

13. Jeremiah O'Sullivan, "Old Ireland and her Monasticism," in Robert McNally, ed.,
Old Ireland (New York, 1965), p. 91.

14. See, e.g., the discussion of Henry Crosse in Michael Walzer, The Revolution of the
Saints (Cambridge, 1965), pp. 207-11.

15. Ibid., pp. 176, 186, 189, 193, passim., on organic imagery and ideas of the family.

FALLACIES OF FACTUAL VERIFICATION

51

think that to prove any previous proposition, which necessarily leads to the
conclusion adverse to them, is taking an unfair advantage? 16

Something like this, without any presumption of rhetorical dirty
dealing, appears in an English textbook on logic which uses a statement
by Arnold Toynbee as a specimen of a petitio. Toynbee wrote, with his
usual obscurity,

We may perhaps take it as having been already demonstrated that an
historian's professed inability to discern any plot, rhythm, or predetermined
pattern is no evidence that blind Samson has actually won his boasted
freedom from the bondage of "Laws of Nature." The presumption is
indeed the opposite; for, when bonds are imperceptible to the wearer
of them, they are likely to prove more difficult to shake off than when
they betray their presence and reveal something of their shape and texture
by clanking and galling. 17

The English logician, E. R. Emmet, interprets this to mean "In
other words:

Invisible bonds are hard to shake off

X cannot see any bonds

Therefore they are hard to shake off

Therefore, the presumption is that they are still there. 18

But Mr. Toynbee is saying something more than this. Leaving out blind
Samson and chains and bonds and clanking and galling, Toynbee is
arguing at very great length in volume 9 that all historians should
generalize intelligently and accurately, and that it is impossible to
generalize intelligently and accurately without generalizing consciously
and deliberately. Therefore, all historians should generalize consciously
and deliberately. This argument may or may not be substantively cor-
rect (I think it is), but it is structurally sound, and can be articulated in
syllogistic form But no sooner did Mr. Toynbee get part of the state-
ment out, in a great jumble of confused imagery, than the logician
pounced.

£»> The fallacy of the prevalent proof makes mass opinion into a
method of verification. This practice has been discovered by cultural
anthropologists among such tribes as the Kuba, for whom history was
whatever the majority declared to be true. 19 If some fearless fieldworker

16. Augustus de Morgan, Formal Logic, 1847 (London, 1926), pp. 296-97.

17. Arnold J. Toynbee, A Study of History, 12 vols. (New York, 1935-61) 9:196.

18. E. R. Emmet, The Use of Reason (London, 1960), p. 226.

19. Jan Vansina, Oral Tradition (Chicago, 1965), pp. 104-5.

52 INQUIRY

were to come among the methodological primitives who inhabit the
history departments of the United States, he would find that similar cus-
toms sometimes prevail. There are at least a few historians who would
make a seminar into a senate and resolve a professional problem by
resorting to a vote. I witnessed one such occasion (circa 1962) as a
student at the Johns Hopkins University. A scholar who was baffled by
a knotty problem of fact literally called for a show of hands to settle the
question. An alienated minority of callow youths in the back of the
room raised both hands and carried the day, in defiance of logic, em-
piricism, and parliamentary procedure.

If the fallacy of the prevalent proof appeared only in this vulgar
form, there would be little to fear from it. But in more subtle shapes,
the same sort of error is widespread. Few scholars have failed to bend,
in some degree, before the collective conceits of their colleagues. Many
have attempted to establish a doubtful question by a phrase such as
"most historians agree . . ."or "it is the consensus of scholarly opinion
that . . ." or "in the judgment of all serious students of this problem. . . ."

A historian has written, for example, "While the role of dope in
damping social unrest in early industrial England has not been exten-
sively investigated, every historian of the period knows that it was com-
mon practice at the time for working mothers to start the habit in the
cradle by dosing their hungry babies on laudanum ('mother's blessing,'
it was called)." 20 This statement is often made, and widely believed.
But it has never, to my knowledge, been established by empirical evi-
dence. The reader should note the hyperbole in the first sentence. When
an historian asserts that "X has not been extensively investigated," he
sometimes means, "I have not investigated X at all."

A fact which every historian knows is not inherently more accurate
than a fact which every schoolboy knows. Nevertheless, the fallacy of the
prevalent proof commonly takes this form — deference to the historio-
graphical majority. It rarely appears in the form of an explicit deference
to popular opinion. But implicitly, popular opinion exerts its power too.
A book much bigger than this one could be crowded with examples. One
will suffice here, for the sake of illustration. Every schoolboy knows, and
most schoolmasters, too, that Mussolini made the trains run on time. But
did he? Ashley Montagu observes that "there was little or no truth in it:
people who lived in Italy between the March on Rome (October 22,
1 922 ) and the execution at Como ( 1 945 ) will bear testimony to the fact
that Italian railroads remained as insouciant as ever with regard to time-

20. Theodore Roszak, "Capsules of Salvation," The Nation, April 8, 1968, p. 470.

FALLACIES OF FACTUAL VERIFICATION

53

tables and actual schedules." 21 And yet, the myth still runs its rounds,
with a regularity that // Duce was unable to bring to his railroads.

$*> The fallacy of the possible proof consists in an attempt to
demonstrate that a factual statement is true or false by establishing the
possibility of its truth or falsity. "One of the great fallacies of evidence,"
a logician has observed, "is the disposition to dwell on the actual pos-
sibility of its being false; a possibility which must exist when it is not
demonstrative. Counsel can bewilder juries in this way till they almost
doubt their own senses." 22 This tactic may indeed prove to be forensically
effective in an Anglo-American court of law, but it never proves a point
at issue. Valid empirical proof requires not merely the establishment of
possibility, but an estimate of probability. Moreover, it demands a bal-
anced estimate of probabilities pro and con. If historians, like lawyers,
must respect the doctrine of reasonable doubt, they must equally be
able to recognize an unreasonable doubt when they see one.

Consider the following example, from the work of an American
historian, Professor J. S. Auerbach. In a recent essay, Auerbach at-
tempted to impeach the authenticity of a statement widely attributed to
Woodrow Wilson. Many texts and treatises assert that, on the night before
Wilson delivered his war message to Congress, he confessed to a
journalist friend that "he'd never been so uncertain of anything in his
life," and that he foresaw among the consequences of his act a "dictated
peace, a victorious peace" imposed upon Germany, and "illiberalism"
imposed upon America.

"Once lead this people into war," the President allegedly said, "and
they'll forget there ever was such a thing as tolerance. To fight you must
be brutal and ruthless, and the spirit of ruthless brutality will enter into
the very fibre of our national life, infecting the Congress, the courts, the
policeman on the beat, the man in the street. ... the Constitution would
not survive it . . . free speech would have to go." 23

Did Wilson actually speak these words? Auerbach says no: "The
sources for the conversation, its content, the circumstances of its publica-
tion, and all that we know about Woodrow Wilson strongly suggest that
the President never spoke the words so frequently attributed to him." 24

21. Ashley Montagu and Edward Darling, The Prevalence of Nonsense (New York,
1967), p. 19.

22. Augustus de Morgan, Formal Logic, 1847, p. 321.

23. John L. Heaton, Cobb of "The World" (New York, 1924), pp. 268-70.

24. J. S. Auerbach, "Woodrow Wilson's 'Prediction' to Frank Cobb: Words Historians
Should Doubt Ever Got Spoken," Journal of American History 54 (1967): 609-17.

54 INQUIRY

But that conclusion rests upon a fallacious method of proof. Auerbach
attempts to demonstrate that Wilson probably did not make his "predic-
tion" by dwelling primarily upon the possibility of error in the evidence.
First, he shows that the earliest published account of the alleged con-
versation was "hearsay evidence, twice removed," which first appeared
in John L. Heaton's biography of the journalist-confidant Frank Cobb.
Heaton heard the story from Maxwell Anderson and Laurence Stallings,
who allegedly heard it from Cobb, who allegedly heard it from Wilson.
Heaton's book came out in 1 924, seven years after the event, three years
after the death of Wilson, and one year after the death of Cobb. A third
of Auerbach's essay is an attempt to impeach the credibility of Anderson
and Stallings in this respect.

Second, Auerbach writes that there is "no evidence whatsoever"
that Frank Cobb visited the White House on the night of April 1, 1917
— the night before Wilson's war message and the night when the con-
versation allegedly took place. 25 He asserts that "it is conceivable that
such a meeting could have passed unrecorded. But it is also conceivable
that the absence of documentation, in conjunction with other evidence,
indicates that the words that presumably were spoken at a nonexistent
meeting on April 2 never were spoken at all." 26

Third, Auerbach argues that the words attributed to Wilson are
suspiciously similar to phrases which did not become common until after
1917, particularly the reference to a "dictated peace."

All of this may seem to establish a plausible case against the authen-
ticity of the alleged conversation. But it does not establish a probable
case. Auerbach's method merely establishes that the conversation might
possibly have been imaginary. There is no presumptive evidence of
probability. Hearsay evidence is not categorically invalid in historical
research, nor is it always inadmissible in a court of law, popular belief
notwithstanding. The "no evidence" of Cobb's visit is precisely what it
says — no evidence — a fallacy of negative proof. The "dictated peace"
argument, on the other hand, commits the fallacy of the partial proof.
The invalidation of a literal phrase in the purported conversation, even
if it were accomplished by Auerbach (it is not), would not suffice to
invalidate the substantive version of the conversation.

There is, however, another approach to the problem. Did Wilson
say similar things to other people at the same time that he allegedly had
his conversation with Frank Cobb? Auerbach says no — the content of the

25. Auerbach quotes from Arthur S. Link's Wilson: Campaigns for Progressivism and
Peace, 1916-1917 (Princeton, 1965), pp. 398-99.

26. Auerbach, p. 612.

FALLACIES OF FACTUAL VERIFICATION

55

Cobb conversation runs "completely counter to the President's words
and deeds." But this is factually false. In the memoirs of Josephus Daniels,
there is an account of a conversation sometime early in 1917, in which
Wilson reportedly said:

There are two reasons why I am resolved to keep our country out of this
war if possible:

1 . If we go to war thousands of young men will lose their lives. I could
not sleep with myself if I do not go to the extreme limit to prevent such
mourning in American homes.

2. Every reform we have won since 1912 will be lost. We have got new
tariff, currency, shipping and trust legislation. These new policies are
not thoroughly set. They will be imperilled or lost if we go to war. We
will be dependent in war upon steel, oil, aluminum, ships and war materials.
They are controlled by Big Business. Undoubtedly many captains of
industry will be patriotic and serve their country, but when the war is
over those whose privileges we have uprooted or started to uproot will
gain control of government and neither you nor I will live to see government
returned to the people. Big Business will be in the saddle. More than that —
Free speech and other rights will be endangered. War is autocratic. 27

This evidence is something less than impeccable. It was published
long after the event, by a man whose memory was less than infallible.
But it serves in some degree to confirm the substantive authenticity of
the Cobb conversation. There is more evidence in interviews which Wilson
held with Ida Tarbell, in transcripts of speeches delivered by Wilson dur-
ing the campaign of 1916, in Wilson's scholarly writings on American
history, and sundry other sources, all of which suggest that Wilson did
agonize most deeply over the coming of war in 1917, did fear the effect
of war upon his program, did believe that war was autocratic and hostile
to freedom, and did desire to preserve civil liberties as he understood
them — though not, perhaps, as Auerbach understood them.

Auerbach did not conduct his inquiry in such a way as to elicit
this factual information. He did not do research in the appropriate
sources, in order to establish the balance of probability by a comparative
analysis of the content of the Cobb conversation in relation to other
relevant evidence for Wilson's thoughts. Instead he confined himself
largely to a discussion of the possibility of error in the evidence at
hand.

The fallacy of the hypostatized proof, as identified and defined
by Perrell F. Payne, "consists in identifying the received theory about

27. Josephus Daniels, The Wilson Era, Years of Peace, 1910-1917 (Chapel Hill, N.C.,
1944), pp. 581-82.

56 INQUIRY

X . . . with X itself, and hence rejecting some variant theory of X on
the grounds that it does not do justice to the nature of A'." 28 In historical
scholarship, this form of error commonly occurs when a historian reifies
a historiographical interpretation and substitutes it for the actual historical
event it allegedly represents, and then rejects contradictory interpreta-
tions or affirms compatible ones.

An example appears in Michael Paul Rogin, The Intellectuals and
McCarthy: The Radical Specter (Cambridge, 1967), in which the
author discusses the Federalists. He hypostatizes Louis Hartz's interpre-
tation of the Federalists — an interpretation which is far off the mark —
and then proceeds to reject all variant interpretations. A few scraps of
allegedly empirical evidence are introduced as reinforcement — scraps and
snippets from Madison and John Adams, who were not representative
Federalists, and indeed scarcely Federalists at all.

The fallacy of the appositive proof is a common but complex
form of empirical error, which consists in an attempt to establish the
existence of a quality in A by contrast with a quality in B — and B is mis-
represented or misunderstood. This is an invidious mistake, which silently
insinuates itself upon the understanding of both the author and his read-
ers, for A is frequently the focus of both the readers' attention and the
author's research, and the erroneous B is bootlegged into the book with-
out warning or conscious control. The result is a little like the old army
shell game, in which many a raw recruit is swindled out of his savings
by being invited to guess which shell covers the pea, and his attention is
diverted from the sergeant's sleight of hand at a crucial moment. But this
fallacy is unlike the shell game in one important way: no swindle is
involved. The perpetrator is the principal victim.

There are many examples. One of them appears in Samuel Eliot
Morison's biography of his ancestor, Harrison Gray Otis. Despite its
many merits, the book tends to be a vindication of Otis and an apology
for the principal acts of his career. Morison is at some pains to show that
his great-grandfather was a statesman and a successful practitioner of
moderate conservative politics. But unfortunately for this interpretation,
Otis was himself the author and abetter of several unstatesmanlike and
spectacularly unsuccessful proposals — most notoriously, the radical in-
novations recommended by the Hartford Convention in 1814. Morison
makes his case not by denying these plain and familiar facts but by con-

28. Perrell F. Payne, Jr., "A Note on a Fallacy," Journal of Philosophy 55 (1958):
124-28.

FALLACIES OF FACTUAL VERIFICATION

57

trasting Otis with a semimythical society of stubborn malcontents called
the Essex Junto, who purportedly possessed great power in New England
politics, and a great desire to smash up the Union after they lost control
of it. Morison suggests that the measure of Otis's moderate statesmanship
was his skill in keeping New England off the rocks and shoals of Essex
Junto extremism. To demonstrate more clearly his ancestor's proficiency
in political navigation, he tries to show that the rocks were dangerous,
and the shoals were treacherous, and the current strong. The more power-
ful and reactionary the Essex Junto appears, the more statesmanlike seems
Otis. By this means, the great political disaster which was the Hartford
Convention is suddenly converted into a shining triumph of unselfish
and successful moderate statesmanship. There is one mistake in all of
this: the fabled Essex Junto is a myth. 29 Morison's error is like Macau-
lay's who is said to have made pygmies of all Europe in order to natter
William III.

A second example is Daniel Boorstin's The Americans: The Colonial
Experience (New York, 1958). Boorstin contrasts the uniqueness of
America with Europe — the "active enterprises of the American" against
the "contemplative tradition of European man"; the intellectual freedom
of Americans against the abstract and sterile system-building of "garret-
spawned illuminati." Boorstin believes that in the eighteenth century
"the best European minds of that age labored to build new-model walls
in which they were to be confined." He lumps together Kant and Hume,
d'Alembert and Diderot, Voltaire and Condorcet, Holbach and Adam
Ferguson in one great musty neo-Thomistic system.

Boorstin's interpretation of the European enlightenment, an inter-
pretation which owes much to Carl Becker's Heavenly City of the Eight-
eenth-Century Philosophers, is simple-minded, uninformed, and seriously
inaccurate. It is fundamental to his interpretation of American unique-
ness, and it is insinuated into the book in such a way that the attention
of the reader (and the author) is deflected from its deficiencies and is
focused instead upon Boorstin's stunning epigrams and appeals to self-
evidence in his discussion of America.

A third example appears in recent works on American Negro
slavery and Negro personality. It is argued by Frank Tannenbaum,
Stanley Elkins, and others that slavery in the United States was uniquely
dehumanizing in its effects on the enslaved, that no other form of slavery
so thoroughly deprived a slave of all the rights and responsibilities of

29. S. E. Morison, The Life and Letters of Harrison Gray Otis, 2 vols. (Boston, 1913).
See also my article, 'The Myth of the Essex Junto," William and Mary Quarterly, 3d
ser. 21 (1964): 191-235.

58 INQUIRY

humanity. There is, I think, a sound and significant element of truth in
this interpretation, but it is distorted by an overdrawn contrast between
Anglo-America and Latin America. Slavery in the latter colonies is
stereotyped, and maybe even a little romanticized, by Elkins, Tannen-
baum, and Klein in their juxtaposition of the two. 30

To condemn this fallacy is not, of course, to condemn the flourish-
ing discipline of comparative history. The most dangerous examples of
the fallacy of appositive proof often appear in works that are not avow-
edly comparative but implicitly and erroneously so. The remedy for
comparative error is not less comparative history but more of it, more
that is explicit and more that is empirical.

The fallacy of misplaced literalism is a form of context error,
which consists in the misconstruction of a statement-in-evidence so that
it carries a literal meaning when a symbolic or hyperbolic or figurative
meaning was intended; or the attribution of a general meaning where a
special or specific one was meant. Messrs. Barzun and Graff comment,
"Misplaced literalism . . . has many forms, and it is particularly insidious
because the reporter must begin by being literal. He must ascertain with
all possible precision what his original text tells him. . . . [But] if he
remains baldly literal and contents himself with quoting extracts, he in-
variably ends by showing his human subject to have been a mass of
contradictions. . . . Misplaced literalism makes a shambles of intellectual
history." 31

John Marshall declared in the Virginia ratifying convention, "We,
Sir, idolize democracy." That statement is surely not to be taken literally.
A glance at the contours of John Marshall's career suggests that he did
not idolize Jeffersonian democracy, or Jacksonian democracy, or indeed
democracy in any accepted meaning. The word "democracy" in his state-
ment is as hyperbolical as the word "idolize." In context, Marshall merely
meant that he and his Federal friends were prepared to acquiesce in a
particular kind of popular election.

Misplaced literalism can also make a shambles of institutional his-
tory. An example comes from the history of industrialization in England.

30. Frank Tannenbaum, Slave and Citizen, 2d ed. (New York, 1963); Stanley Elkins,
Slavery: A Problem in American Institutional and Intellectual Life, 2d ed. (New York,
1963); Herbert S. Klein, Slavery in the Americas: A Comparative Study of Cuba and
Virginia (Chicago, 1967).

31. Jacques Barzun and Henry Graff, The Modern Researcher, 2d ed. (New York, 1962),
pp. 127-28.

FALLACIES OF FACTUAL VERIFICATION

59

In a work published in 1619, there appears the following description of
a factory which was allegedly visited by Henry VIII:

Within one room being large and long

There stood two hundred looms full strong

Two hundred men, the truth is so,

Wrought in these looms all in a row.

By every one a pretty boy

Sat making quils with mickle joy.

And in another place hard by

An hundred women merrily,

Were carding hard with joyful cheer

Who singing sat with voices clear.

And in a chamber close beside

Two hundred maidens did abide

In petticoats of Stamell red

And milk white kerchers on their head

These pretty maids did never lin

But in that place all day did spin.

This source has been literally interpreted by historians as evidence
of the existence of factories in the sixteenth century. The alleged owner,
Jack of Newbury, has been identified as a great textile magnate. But
Peter Laslett observes that "this Jack of Newbury was as much of a myth
as Jack and the Beanstalk." 32 The poem appeared in a novel which is as
suitable to a literal interpretation by a student of sixteenth-century Eng-
land as Gargantua and Pantagruel might be for a student of sixteenth-
century France. Imagine what might happen if all evidence for the latter
were destroyed except the works of Rabelais.

Still another specimen of misplaced literalism occurs in Darrett B.
Rutman's monograph Winthrop's Boston: Portrait of a Puritan Town,
1 630-1649 (Chapel Hill, 1965). The author argues that the founders
of Boston wished to establish a "medieval monastery," in which society
would be "perfectly united in thought, speech, judgment, and, above all,
God's holy love." Much of Rutman's thesis rests upon a lay sermon which
John Winthrop delivered on board the ship Arbella in passage to New
England. Rutman quotes excerpts from this source as literal evidence of
Winthrop's expectations. The most famous line — "Wee shall be as a
Citty upon a Hill" — brings forth the following comment: "Winthrop's
expression was much more than a literary conceit borrowed from the
Gospel of Matthew. It reflected the core of his thinking about the society
he and his fellows intended to establish." 33

32. Laslett, The World We Have Lost (New York, 1965), pp. 152-53.

33. P. 4

60

INQUIRY

But did it, exactly? Rutman makes a sermon into a social blueprint,
a hortatory message into a measure of both a man's specific purpose and
his general sense of prospects and possibilities. Imagine a football team
which is losing, 42-3, at the end of the first half. In the locker room,
the coach says, "Let's go, you guys: We need six quick touchdowns:
And I know that you will do it!" Rutman's method would convert this
expression of hope into an indication of what the coach really plans and
expects. More likely, the coach had already called his wife and said, "I'll
be home early. No bonfire tonight!"

A reviewer of Rutman's book complains, "My only regret is that
the author has pursued what is essentially [sic] a literary device to a
point where it seems to affect his assessment of the Puritans. In taking
an exhortation as a statement and in using it as a reference point through-
out the book, he makes Winthrop and the other founders appear more
naive than they were." 34

Misplaced literalism does not appear only in the interpretation of
literary evidence. The same mistake is easily made in the use of other
kinds of sources, too. Consider, for example, the party in the White
House following Andrew Jackson's inauguration. The goings-on were
described by an eyewitness, Mrs. Margaret Bayard Smith:

What a scene did we witness! The Majesty of the People had disappeared,
and a rabble, a mob, of boys, negros, women, children, scrambling fighting,
romping. What a pity what a pity! No arrangements had been made no
police officers placed on duty and the whole house had been inundated
by the rabble mob. We came too late. The President, after having been
literally nearly pressed to death and almost suffocated and torn to pieces
by the people in their eagerness to shake hands with Old Hickory, had
retreated through the back way or south front and had escaped to his
lodgings at Gadsby's. Cut glass and china to the amount of several thousand
dollars had been broken in the struggle to get the refreshments, punch
and other articles had been carried out in tubs and buckets, but had it
been in hogsheads it would have been insufficient, ice-creams, and cake
and lemonade, for 20,000 people, for it is said that number were there,
tho' I think the estimate exaggerated. Ladies fainted, men were seen with
bloody noses and such a scene of confusion took place as is impossible to
describe — those who got in could not get out by the door again, but had
to scramble out of the windows. At one time, the President who had retreated
and retreated until he was pressed against the wall, could only be secured
by a number of gentlemen forming round him and making a kind of barrier
with their bodies, and the pressure was so great that Col Bomford who was
one said at one time he was afraid they should have been pushed down,
or on the President. It was then the windows were thrown open, and the
torrent found an outlet, which otherwise might have proved fatal. . . . The

34. Edmund S. Morgan in American Historical Review 71 (1966): 664.

FALLACIES OF FACTUAL VERIFICATION

61

noisy and disorderly rabble in the President's House brought to mind
descriptions I had read, of the mobs in the Tuileries and at Versailles, I
expect to hear the carpets and furniture are ruined, the streets were muddy,
and the guests all went thither on foot. 35

This little "scene of confusion" has often been used as a key to
that larger scene of confusion called the Jacksonian era — as apparent
evidence of the rise of the Common Man, whoever he may have been.
Now, let us assume that things happened precisely as Mrs. Smith re-
ported them — a doubtful assumption, for Mrs. Smith's descriptions of
things Jacksonian were a little like her friend Mrs. Bomford's inaugura-
tion gown, deep scarlet, richly trimmed with embroidery. But let us
assume it to be correct. Even so, it is not a literal proof of political equal-
ity in the Jacksonian era. There were similar goings on at the coronations
of medieval kings. At the coronation banquet of Charles VI of France
in 1380, "the throng of spectators, guests and servants was such that the
constable and the marshal of Sancerre had to serve up the dishes on
horseback." At the coronation of Henry VI of England, in 1431, "the
people force their way at daybreak into the great hall where the feast
was to take place, 'some to look on, others to regale themselves, others
to pilfer or to steal victuals or other things.' " During the "inauguration"
of Louis XI, it is said that the disorder was such that "the princes of the
blood were nearly squeezed to death in their seats of honour." 36

To assume that Jackson's inaugural reception was a literal reflection
of the Jacksonian movement itself is surely a fallacy.

The fallacy of misplaced precision is an empirical statement
which is made precise beyond the practical limits of accuracy. One
fanatical quantifier in the sixteenth century was curious to know the
weight of a stone cannon ball, 10% inches in diameter. There was, of
course, considerable variation in the size and weight of stone cannon
balls, even those prepared for a single specific gun. And there was un-
doubtedly some small change in the weight of any single shot, as it was
rubbed and chipped by careless gunners, etc. But our inquirer figured
the weight of a 10% inch ball as "61 lb., 1 oz., 2 drams, 1 scruple and

685644

15 1414944 grains," thus attempting to ascertain the weight of a various
and variable object within a millionth part of the weight of a grain of
barley. The modern scholar who reported this episode commented that

35. Margaret Bayard Smith to Jane Kirkpatrick, March 11, 1829, as published in The
First Forty Years of Washington Society (New York, 1906), pp. 295-96.

36. Johan Huizinga, The Waning of the Middle Ages (Garden City, N.Y., 1954), p. 50.

62 INQUIRY

it showed "a kind of theoretical meticulousness which is quite medieval
in flavor." 37 But it is one which we shall see more of in modern historiog-
raphy as quantification becomes increasingly fashionable.

£*> "Prove everything," saith St. Paul, "and hold fast to that
which is good." Historians are likely to agree in principle, but not in
practice. Specific canons of historical proof are neither widely observed
nor generally agreed upon. There is no historiographical Wigmore,
Stephen, or Thayer and no body of precedents which is recognized as a
reliable guide.

But there are, I think a few simple rules of thumb, which are the
inversion of the fallacies discussed in this chapter. First, as Wigmore has
exhaustively argued, sound evidence consists in the establishment of a
satisfactory relationship between the factum probandum, or the proposi-
tion to be proved, and the factum probans, or the material which is of-
fered as proof. That is sufficiently obvious. But it is not so obvious to
many scholars that the criteria for a satisfactory factum probans depend
in large degree upon the nature of the factum probandum. 3 * This is a
pedantical way of saying that every fact in history is an answer to a ques-
tion, and that evidence which is useful and true and sufficient in answer
to question B may be false and useless in answer to question A. A his-
torian must not merely get the facts right. He must get the right facts
right. From this a simple rule of relevance may be deduced: historical
evidence must be a direct answer to the question asked and not to some
other question.

Secondly, an historian must not merely provide good relevant evi-
dence but the best relevant evidence. And the best relevant evidence, all
things being equal, is evidence which is most nearly immediate to the
event itself. The very best evidence, of course, is the event itself, and
then the authentic remains of the event, and then direct observations,
etc. We shall call this the rule of immediacy.

Third, evidence must always be affirmative. Negative evidence is
a contradiction in terms — it is no evidence at all. The nonexistence of an
object is established not by nonexistent evidence but by affirmative evid-
ence of the fact that it did not, or could not exist, as, for example, in a
low coefficient of correlation, in the case of a comparative analysis. In
the case of a journalist who did not in fact call upon a President, proof

37. Michael Lewis, "Armada Guns: A Comparative Study of English and Spanish
Armaments. Section V. The Guns of the Spanish Fleet, 1588," The Mariner's Mirror
29 (1943): 3-39, 8 n.

38. John Henry Wigmore, A Treatise on the Anglo-American System of Evidence in
Trials at Common Law. . . , 3d ed., 10 vols. (Boston, 1940), 1, sect. 2.

FALLACIES OF FACTUAL VERIFICATION 63

requires affirmative evidence from the journalist or the President, or
evidence that the journalist was somewhere else at the time, or evidence
that the president's time was entirely taken up with other things. If proof
of this sort cannot be found, then the point cannot be proved, and a his-
torian must candidly accept uncertainty. This is the rule of affirmation.

Fourth, the burden of proof, for any historical assertion, always
rests upon its author. Not his critics, not his readers, not his graduate
students, not the next generation. Let us call this the rule of respon-
sibility.

Fifth, all inferences from empirical evidence are probabilistic. It is
not, therefore, sufficient to demonstrate merely that A was possibly the
case. A historian must determine, as best he can, the probability of A
in relation to the probability of alternatives. In the same fashion he
cannot disprove A by demonstrating that not-A was possible, but only
by demonstrating that noi-A was more probable than A . This is the rule
of probability.

Sixth, the meaning of any empirical statement depends upon the
context from which it is taken. No historical statement-in-evidence floats
freely outside of time and space. None applies abstractly and universally.
The statement that a Norman army defeated a Saxon army at Hastings
in 1066 is meaningless without reference to a map of England, and also
to our calendar. For a Moslem, the same event has the different date
of 459. The relativity of a date to a calendrical frame of reference is
self-evident. But the contextual relativity of many other "facts" is not.

Seventh, an empirical statement must not be more precise than its
evidence warrants. And degrees of precision, of course, vary greatly
from one piece of evidence to another. Someday, historians may take a
leaf from statisticians and specify the quantitative "significance" of their
statements. In the meantime, the same effect can be accomplished by
nuances of language. We shall call this the rule of precision.

These seven rules of thumb are ones that good historians feel in
their bones and apply without thinking. If such rules are raised to the
level of consciousness, practices might hopefully improve in at least
some small degree. The factual errors which academic historians make
today are rarely deliberate. The real danger is not that a scholar will
delude his readers, but that he will delude himself.

CHAPTER III

FALLACIES OF FACTUAL
SIGNIFICANCE

No one reads or writes history in a fit of total absent-
mindedness, though a fair amount of history has been
written by people whose minds seem in part to have
been on other things.

— G. R. Elton

To write history, or even to read it, is to be endlessly engaged in a process
of selection. No part of the job is more difficult or more important, and
yet no part has been studied with less system, or practiced with less
method. Many facts are called, but few are consciously chosen, on ex-
plicit and rational criteria of factual significance.

Historians have, I think, deliberately resisted the refinement and
rationalization of this aspect of their task. For some of them, the idea
of a chosen fact is hateful in itself. Others think that the process of
selection is not merely an unknown but a mystery. Both of these be-
liefs are utterly mistaken. The process of selection can and must be
clarified if history is to develop beyond its present condition. Criteria of
factual significance can and must be specified, or else historians will be
running through records like rats in a maze, without even a rudimentery
notion of the nature of their predicament. 1

In every historical inquiry, the process of factual selection operates
in several separate ways. One species of selection, commonly called
"sampling," consists in the selection of representative facts of a certain
predetermined kind, within a closed universe of investigation. This un-
dertaking, difficult as it often is in practice, presents comparatively few

1. An important beginning has been made by two able philosophers of history: Arthur
C. Danto, Analytical Philosophy of History (Cambridge, 1965), pp. 1-16, 132-39; and
Morton White, The Foundations of Historical Knowledge (New York, 1965), pp. 237-
270.

FALLACIES OF FACTUAL SIGNIFICANCE

65

conceptual problems. We shall consider it in a later chapter. A much
tougher problem is presented by the complex selection process that
precedes sampling — namely, the selection of the kinds of facts to be
sampled. This involves the determination of fundamental criteria of
factual significance.

Significance is the sort of word which sends shudders down the
spine of every working historian. But if he turns his back upon the
problem, he will operate upon covert criteria of factual significance,
which may be incompatible with his own objectives. "He who is deficient
in the art of selection," Macaulay wrote, "may, by showing nothing but
the truth, produce the effect of the grossest falsehood." 2 And he who is
deficient in his ideas of significance is sure to be equally so in the art of
selection.

There are a few common fallacies in this respect, a few criteria
of factual significance which are inconsistent with empiricism in history.
In each of the following instances, the fallacy consists not in the criteria
themselves but rather in an attempt to combine them with the methods
and objects of empirical inquiry. Every historian, of course, has the
inalienable right to do any kind of history that pleases him — a constitu-
tional right to go wrong in his own way. But no historian is liberated
from the logical consequences of his own assumptions. The object of
this chapter is not to condemn certain deviations from narrow academic
norms. It is rather to demonstrate that all historians operate upon certain
criteria of factual significance in their work, and that those criteria must
be aligned with their own purposes, and methods.

The holist fallacy is the mistaken idea that a historian should
select significant details from a sense of the whole thing. This method
seems plausible at first sight. But it would prevent a historian from
knowing anything until he knows everything, which is absurd and im-
possible. His evidence is always incomplete, his perspective is always
limited, and the thing itself is a vast expanding universe of particular
events, about which an infinite number of facts or true statements can be
discovered.

An extreme example of holism is Hegel's The Philosophy of History,
which is clearly and cogently discussed by Bertrand Russell:

The view of Hegel and of many other philosophers, is that the character
of any portion of the universe is so profoundly affected by its relations
to the other parts and to the whole, that no true statement can be made

2. Thomas Babington Macaulay, "History," Complete Writings, 20 vols. (Boston,
1900), 11:245.

66

INQUIRY

about any part except to assign its place in the whole. Thus there can be
only one true statement; there is no truth except the whole truth. . . . Now
this is all very well, but it is open to an initial objection. If the above
argument were sound, how could knowledge ever begin? I know numbers
of propositions of the form "A is the father of B," but I do not know
the whole universe. If all knowledge were knowledge of the universe as
a whole there would be no knowledge. This is enough to make us suspect
a mistake somewhere. 3

A historian who swears to tell nothing but the whole truth, would
thereby take a vow of eternal silence. A researcher who promises to
find the whole secret for himself condemns himself to perpetual failure.
The whole truth, at any stage of an inquiry, is an ideal that ought to
be abolished from historiography, for it cannot ever be attained. His-
torians are bound to tell the best and biggest truths they can discover,
but these truths are very different from the whole truth, which does
not and cannot exist. A scholar who seeks the whole truth is on a road
which can only end in the intellectual suicide of relativism, or else
in that condition of methodological anomie which characterizes so
many of my colleagues.

Georg Wilhelm Friedrich Hegel, poor twisted Teutonic soul that
he was, is an easy mark for a methodologist. Most of the fallacies in
this book could be illustrated by his arguments. But there are many
other examples of the holist fallacy, which is an exceedingly common
form of error. All metahistorians, by definition, are guilty of this mis-
take — Toynbee, Spengler, Sorokin, Marx, Comte, Kant, Condorcet,
Vico — and others who have tried to discover the "meaning" of the
whole past. 4

The extravagant holism of metahistory is held in deserved con-
tempt by most working historians. But the same form of error also
comes in smaller sizes. Many a monographer has set his sights upon
the whole of a little subject, like William James's Adirondack hunter
who shot a bear by aiming, not at his eye or heart, but "at him gen-
erally." 6

3. Bertrand Russell, A History of Western Philosophy (New York, 1945), pp. 743, 745.

4. Arthur Danto reaches the same conclusion by a different argument. He argues that
the significance of events is always, in part, dependent upon later events; and therefore,
that the significance of past events is partly dependent upon future events; and thus, that
substantive philosophers of history are condemned to failure in their quest for the whole
truth, for their method would require a "history of events before the events themselves
have happened" (Analytical Philosophy of History, p. 14). This is not to say that the
past can change, for events happened in the way that they happened, and not in any
other way. But facts, or true statements about past events, can and will change, as
other events occur.

5. William James, The Principles of Psychology, 2 vols. (New York, 1890), 2:333-34.

FALLACIES OF FACTUAL SIGNIFICANCE

67

A classical example is Macaulay, whose essay on history is a
disquisition on selection and significance. But his criteria of signifi-
cance are holistic. "No history," he knew, "can present us with the
whole truth; but those are the best . . . histories which exhibit such
parts of the truth as most nearly produce the effect of the whole. . . . The
perfect historian is he in whose work the character and spirit of an
age is exhibited in miniature." 6

But if the whole of history is unattainable, so also is the whole
of any historical age. Macaulay himself noted, inconsistently, that
"if history were written thus, the Bodleian library would not contain
the occurrences of a week." 7 And if the whole of an "age," or any such
part of history can never be known, it can never be exhibited in minia-
ture.

The idea is absurd, and yet it persists. Many modern historians
continue to dig themselves into wholes, large and little, and it isn't
easy to root them out again. Professor G. R. Elton, in a historiographical
how-to-do-it manual, stubbornly insists that "all good historical writ-
ing is universal history in the sense that it remembers the universal while
dealing with a part of it." 8 Professor Jacques Barzun, in his romantic
way, writes blissfully of "the intelligibility of the whole.'" The holistic
fallacy has been much invigorated by the fashionable idea of cultural
history; many scholars flatter themselves, as Barzun does, that they have
discovered in "culture" the whole secret for themselves.

The great unesco history project, which is presently in progress,
is a serious attempt to synthesize the sum of all human memories and
to recapture the "past in its entirety." The second volume of this
Quixotic venture promises to tell the whole truth about the ancient
world, from 1200 B.C. to a.d. 500. 10 It was prepared by thirty-seven
contributors and consultants. The result is a catastrophe on an appro-
priately monumental scale. A critic comments, "Rarely, if ever, can
so many learned men have labored so long on a history to so little
purpose." 11 The book is as bland and trivial as if the members of the
Security Council had collaborated on a treatise of political theory. A
project designed to explain everything ends, predictably, by explaining
nearly nothing.

6. Macaulay, "History," pp. 244^45, 280.

7. Ibid., p. 244.

8. G. R. Elton, The Practice of History (New York, 1967), p. 16.

9. Jacques Barzun, "Cultural History as a Synthesis," in Fritz Stem, ed., The Varieties of
History (New York, 1956), p. 393.

10. Luigi Pareti, et al., The Ancient World, 1200 BC to AD 500: History of Mankind,
Cultural and Scientific Development (New York, 1965).

11. Laurence Lee Howe, The American Historical Review 71 (1966): 520.

68 INQUIRY

The holist fallacy may be developing in a fresh new form today,
thanks to the electronic revolution. An economist, Kenneth E. Bould-
ing, seriously proposes that the whole of recorded history should be
programed on one gigantic IBM machine. "One visualizes a computer
on which the totality of recorded history has been coded and from
which then samples can be taken, relations perceived, discrepancies
identified, and continuously new questions asked and gaps in the data
discovered. The problem of coding would indeed be a difficult one,
but it is surely not beyond the capacity of the historian's intelligence." 12

Some recorded history is presently being encoded into a computer-
readable form, in various data banks throughout the Western world.
We shall all gain much from these immense projects. But nobody will
ever program the whole of recorded history into a computer — merely
a few selected kinds of material. Enormous though they may be, their
selection will require the same careful and explicit criteria of signifi-
cance which smaller projects now demand.

The fallacy of essences begins with the old idea that every-
thing has something deep inside it called an essence, some profound
inner core of reality. According to this view, facts about a man, a
nation, an age, a generation, a culture, an ideology, or an institution
are significant in the degree to which they display the essence of the
entity in question.

This most durable of secular superstitions is not susceptible to
reasoned refutation. The existence of essences, like the existence of
ghosts, cannot be disproved by any rational method. But it is possible
to demonstrate that a belief in essences, like a belief in ghosts, involves an
empiricist in certain difficulties. This has been done at some length by
Karl Popper, who persuasively suggests that the progress of empirical
knowledge requires, not a search for essences, which cannot be found by
any empirical method, but rather a search for patterns of external be-
havior. 13 The essentialist's significant facts are not windows through which
an observer may peek at the inner reality of things but mirrors in which
he sees his own a priori assumptions reflected.

The fallacy of essences is, nevertheless, very common in histori-
cal writing. It is psychologically gratifying, for it supplies a sense of
completeness and it encourages a sense of certainty. But these are illu-
sions which an empiricist must learn to live without. Popper believes

12. History and Theory 7 (1968): 90.

13. Karl Popper, The Open Society and Its Enemies, 2 vols. (New York, 1963), 1,
chap. 3, pt. 6; The Poverty of Historicism (Boston, 1957), chap. 1, pt. 10.

FALLACIES OF FACTUAL SIGNIFICANCE

69

that natural scientists have learned to live without them. He correctly
complains that "the problems of the social sciences, on the other hand,
are still for the most part treated by essentialist methods. This is, in my
opinion, one of the main reasons for their backwardness." 14

Essentialism is, I think, closely related to holism, for knowledge of
the "essence" of a thing implies knowledge of the whole thing, which
is impossible. Consider, for example, the case of Sir Lewis Namier, a
great historian and a gross Tory (the organic bias of classical con-
servatism perhaps explains the particular susceptibility of conservatives
to holism and essentialism). Sir Lewis was no enemy of chosenness in
either facts or people. He was, indeed, a confirmed Zionist in both
respects. The role of selection in historical writing and the importance
of "significant detail" were explicitly acknowledged in his works.

But Namier so conceptualized the selection process as to under-
cut its very existence and at the same time to diminish the empirical
aspect of historical knowledge.

The function of the historian [he wrote] is akin to that of the painter
and not of the photographic camera: to discover and set forth, to single
out and stress that which is of the nature of the thing, and not to reproduce
indiscriminately all that meets the eye. To distinguish a tree you look
at its shape, its bark and leaf; counting and measuring its branches would
get you nowhere. Similarly what matters in history is the great outline
and the significant detail; what must be avoided is the deadly morass of
irrelevant narrative. History is therefore necessarily subjective and indi-
vidual. 15

The essentialist fallacy appears in Namier's notion of "the nature
of the thing," which, if it means anything at all, must mean some intui-
tive sense of the inner essence of the thing. The holist fallacy appears
in the "great outline," which is a holistic greatness, because it is
general and all-inclusive. One must surely agree with Namier as to
"significant detail," but not as to his standard of significance, which is
not a method of selection at all but a pious and stubborn hope, against
all reason and experience, that a historian will be able to operate
intuitively without one.

A second example comes from the opposite end of the political
spectrum and suggests a relationship between the fallacy of essences
and another form of error, which we shall come to in this chapter —
the antiquantitative fallacy. Barrington Moore, in Social Origins of
Dictatorship and Democracy (Boston, 1966), attacks all statistics
that contradict his radical presuppositions. "Statistics,",, he writes, "are

14. Popper, The Open Society and lis Enemies, 1:33.

15. Sir Lewis Namier, "History," Avenues of History (London, 1952), p. 8.

70 INQUIRY

misleading traps for the unwary reader when they abstract from the
essence of the situation the whole structural context in which social
osmosis takes place. As statistics are fashionable now, it is worthwhile
stressing this point." 16 I shall have something more to say about this
statement later in this chapter. Here, one might note that Moore has
things backwards. It is not external facts which are abstracted from
essences by observers, but essences from external facts. Moore's under-
standing of the essence of historical events is a radical vision of the
progress of humanity toward some egalitarian Utopia. All external
facts and statistics which sustain this vision are significant for Moore;
all those which contradict it are not. Can anything be more absurd?

Namier and Moore are not alone in their essentialism. In between,
many historians have made the same error. Morton White suggests that
any historian who argues that "Europe was essentially characterized
by a Renaissance at one time, and at another by an Enlightenment,
may think along similar lines." 17 And similarly, "even the most cau-
tious and unmetaphysical of historians are prepared to speak in some
such vein of the 'line' that the main stream of American 'actuality' has
followed." 18 White is referring to the "Jefferson-Jackson-Franklin Roo-
sevelt" line and the "Federalist-Whig-Republican line" which many his-
torians have identified as the "essence" of American history.

The fallacy of essences is tempting to historians, because many of
them begin with an article of faith (which I happen to share) that
history happened in the way that it happened and not in any other
way. But it does not follow from this premise that there is one "essential"
inner reality, which can be hunted and found. There are many factual
patterns — an infinite number of them — which can be superimposed
upon past events. A historian's task is to find patterns which are more
relevant to his problems, and more accurate and more comprehensive
than others, but he cannot hope to find that "essential" pattern, any
more than he can hope to know all of history, and to know it ob-
jectively. Here again, the progress of historical science depends squarely
upon a sense of the possible.

The prodigious fallacy mistakes sensation for significance.
It is the erroneous idea that a historian's task is to describe portents
and prodigies, and events marvelous, stupendous, fantastic, extraor-
dinary, wonderful, superlative, astonishing, and monstrous — and further,

16. P. 37; italics added.

17. White, The Foundations of Historical Knowledge, p. 244.

18. Ibid., p. 244.

FALLACIES OF FACTUAL SIGNIFICANCE

71

that the more marvelous, stupendous, etc., an event is, the more his-
toric and eventful it becomes. This absurd standard of significance is
older than history itself. Herodotus, the putative "father of history"
(grandfather might be more correct), composed his great work on the
explicit assumption that a historian should entertain his readers with
tales of true "wonders." The result was a perpetual tension in his work
between that which was true and that which was wonderful, a tension
which was profoundly dysfunctional to the historicity of his interpreta-
tion. 19

Today, this form of error is widely disseminated by the mass
media. Journalists use the word "historic" to describe earthquakes,
hurricanes, five-alarm fires, floods, state funerals, typhoons, train wrecks,
quintuplets, transatlantic canoe trips, and other curiosities or catas-
trophes which are chiefly remarkable for the fortunate fact that they
rarely occur, or for the unfortunate fact that they occur at all.

War reportage from Vietnam is studded with these fallacies,
which are known to their perpetrators as "the left-handed battalion
commander syndrome." A journalist explains: "It was a function of
journalistic desperation to differentiate one military operation from
another. An enormous effort was made to establish a 'first' or 'most' or
'least' in the lead of a newspaper article. It was surmised that the
classic lead for the nonevent of a fruitless operation [sic] would be that
a left-handed battalion commander, 'for the first time,' led it into bat-
tle." 20

One expects to see this sort of thing on the sports page, where
history is made by an infield triple, or a one-armed shortstop, or the
most Texas Leaguers in a ten-inning game. But somehow it always
comes as a surprise to find it on the editorial page of The New York
Times, where the defection of Stalin's daughter was played up for many
months in 1967 as one of the "historic" events of the century. And it
is even more disconcerting to hear a good journalist-turned-historian,
William Manchester, saying that "the abdication of Edward VII [sic]
. . . was the greatest story since the Resurrection." 21

In the genre called popular history, many prodigious fallacies
are perpetrated by historians who appear to have mistaken their muse
for Miss Dorothy Kilgallen, and their mission for the titillation of
illiterate thrill-seekers. Almost any given volume in Doubleday and
Company's disastrous Mainstream of America series can serve as an

19. The History of Herodotus, trans. George Rawlinson, ed. Manuel Komroff (New
York, 1956), pp. 36, 134, 149, 187, 213, 267.

20. Ward S. Just, To What End? (Boston, 1968), p. 15.

21. Quoted in John Corry, The Manchester Affair (New York, 1967), p. 86.

72 INQUIRY

example. A particularly serious specimen is Paul I. Wellman's The
House Divides (Garden City, N.Y., 1966), which purports to be a
serious popular history of the United States from the end of the War
of 1812 to the start of the Civil War. Wellman appears to have be-
lieved that the best way to bring history to life is to crowd it with
as many violent deaths as possible. In his fourth chapter, ostensibly
about the early career of Andrew Jackson, I counted 1,452 violent
killings. Many are described in graphic detail. A few are described
twice. This butcher's bill for Chapter 4, it should be noted, does not
include the battle of New Orleans, which has Chapter 1 all to itself —
an entire celebratory chapter dedicated to the sublimity of that epic
slaughter. 22

Political history, in Wellman's book, is retailed in a sequence of
anecdotes and apocrypha — the false fable about Stephen Van Rens-
selaer and the election of 1824 (p. 68); the alleged vote of an invalid
in Switzerland County, Indiana, which "made Texas a state and brought
on the Mexican War" (p. 213); the Peggy Eaton affair, which is in-
terpreted as a "trigger" of the Civil War; and other assorted marvels,
mostly false, about Abraham Lincoln's first sight of slavery (p. 286);
the defense of the Alamo (p. 157); Daniel Webster's seventh-of -March
speech (p. 330); and the underground railroad (p. 236).

Economic history is largely left out of sight. The only extended
discussion of a manufacturing process is a recipe for "Injun Whiskey."
As for social, intellectual, cultural, and educational history, Jim Bowie
receives more space than all major figures in these areas combined.
Chief Justice John Marshall gets seven lines; James Marshall, the
mechanic who found gold in the American River, gets sixty-three.

Wellman's book is not much worse than many others in the same
series, which has been widely distributed by Doubleday, and often,
I fear, urged upon innocent children by well-meaning librarians and
school teachers. 23

22. For violence, see in addition to chap. 4, pp. 1-12, 15-21, 26, 30-36, 37-55, 74, 78,
80,99, 100, 118-19, 130, 135, 144-45, 148-63, 184-86, 188, 190-92, 194, 196, 199,
201, 202, 207, 216-80, 288-89, 295, 299, 300-1, 313-14, 345-49, 350-53, 354-62,
367, 378-79, 380-81, 388, 389, 400, 415-17. For sex, see pp. 25, 33, 60, 65, 73, 75, 76,
78-79, 86, 87, 91-92, 106-7, 120-23, 161, 206, 248-49, 253, 291, 294, 296-98,
352, 377-78.

23. A few Mainstream of America volumes are exceptional. They tend to be books
avowedly on military subjects, such as Bruce Catton's This Hallowed Ground, C. S.
Forester's Age of Fighting Sail, and Edward Hamilton's French and Indian Wars. Vol-
umes of colonial history by Harold Lamb and Marion Starkey are at least accurate.
And David Lavender's Land of Giants is a good history of the Pacific Northwest. For
deficiencies in the others, the reader might consult critical reviews by competent
authorities. I recommend for Hodding Carter's The Angry Scar the review in the Journal

FALLACIES OF FACTUAL SIGNIFICANCE

73

But popular historians have no monopoly on prodigious fallacies.
Academic historians tend to commit them more delicately but not less
frequently. There are examples even in the Oxford History of England
series, magisterially edited by Sir George Clark. The most recent volume
is English History, 1914-1945, by A. J. P. Taylor (Oxford, 1965).
The reader is invited to locate Taylor's standards of significance, in
the following biographical notes:

J George V (1865-1936), second son of Edward VII: married Princess
Mary of Teck, 1893; king, 1910-1936; changed name of royal family from
Saxe-Coburg to Windsor, 1917; his trousers were creased at the sides,
not front and back [p. 2n].

2 David Lloyd George (1863-1945) : educ. Church school; Liberal M.P.,
1890-1945; chancellor of the exchequer, 1908-15; minister of munitions,

1915- 16; secretary for war, 1916; prime minister, 1916-22; leader of the
Liberal party, 1926-31; cr. Earl Lloyd-George, 1945. A master of improvised
speech and of improvised policies. Though he was dangerous to most
women, he gave his heart to few. After leaving office, he farmed ambitiously,
though unprofitably, and propagated the 'Lloyd-George' raspberry He
disliked his correct surname, 'George,' and imposed 'Lloyd George' on
contemporaries and posterity [p. 5n].

■Asquith was the first prime minister since the younger Pitt who is said
to have been manifestly the worse for drink when on the Treasury Bench.
George Robey was uncomfortably near the mark when he sang:
Mr. Asquith says in a manner sweet and calm:
Another drink won't do us any harm [p. 15n].

"Walter Long (1854-1924): educ. Harrow and Oxford; M.P., 1882-
1921; president of local government board, 1915-16; colonial secretary,

1916- 18; first lord of the admiralty, 1918-21; cr. Viscount, 1921. Chesterton
wrote (incorrectly) :

of Southern History 25 (1959) : 400-402; for Clifford Dowdey's The Land They Fought
For, ibid., 21 (1955): 541-42; for John Dos Passos' Men Who Made the Nation,
William and Mary Quarterly, 3d ser. 15 (1958): 120-21; for Bruce Lancaster's From
Lexington to Liberty, ibid., 12 (1955): 665-67; for Irving Stone's Men to Match My
Mountains, see Pacific Historical Review 26 (1957): 79-80; and for Paul Wellman's
earlier volume, Glory, God, and Gold, ibid., 24 (1955): 80-81. Stewart Holbrook's
The Age of the Moguls is reviewed in The Saturday Review of Literature, Oct. 10,
1953, and bis Dreamers of the American Dream, in ibid., Oct. 26, 1957. I did not find
good reviews of Kenneth Davis's The Experience of War, which promiscuously mixes
fiction and fact; or Saunders Redding's Lonesome Road, which is a moving human
document, but not a sound or reliable history of the Negro in America.

Such works as these are not commonly reviewed in learned journals, which is un-
fortunate. If The American Historical Review, crowded though it is, were to make an
effort to stimulate careful critiques of popular histories, and then to distribute those
critiques to librarians and public school teachers, the quality of pop history would soon
improve. If the reader will forgive a hackneyed phrase, pop history is much too serious
a business to be left to pop historians and businessmen.

74 INQUIRY

Walter, beware! scorn not the gathering throng . . .
It suffers, but it cannot suffer Long [p. 40n].

There are many other "true wonders" in Taylor's notes, and more
in his text, and a few even in his index. Some are more wondrous than
true. 24 Together, they suggest that the youthful spirit of Herodotus has
revived in the ripe old age of English academic history. Professor Taylor
has, I think, hunted down his smoking-room stories in the same Herodo-
tean spirit which caused that ancient author to journey all the way from
Halicarnasus to Arabia to investigate reports of winged serpents, and
to regale his readers with strange tales of giants and dwarfs, men
with horns and oxen without them, savage tribes who worshiped their
parents and others who ate them. 25

The furtive fallacy is the erroneous idea that facts of special
significance are dark and dirty things and that history itself is a story
of causes mostly insidious and results mostly invidious. It begins with
the premise that reality is a sordid, secret thing; and that history hap-
pens on the back stairs a little after midnight, or else in a smoke-filled
room, or a perfumed boudoir, or an executive penthouse or somewhere
in the inner sanctum of the Vatican, or the Kremlin, or the Reich
Chancellery, or the Pentagon. It is something more, and something other
than merely a conspiracy theory, though that form of causal reduction
is a common component. The furtive fallacy is a more profound error,
which combines a naive epistemological assumption that things are

24. Henry Pelling, 'Taylor's England," Past and Present 33 (1966): 149-58.

25. Taylor is not the only academician who has slipped into his "Herodotage," as
Malinowski called it Occasional lapses are apparent in the work of all scholars — even
duly licensed American Ph.D.s! Eugene L. Cox, in a fine book called The Green Count
of Savoy (Princeton, 1967), was unable to resist the following marvel:

"The conflicting claims of the Pope and Bernabo Visconti finally led to Bernabo's
excommunication. Milanese chroniclers recount that the Lord of Milan was sitting with
some of his men on a bridge over the Lambro when the papal envoys approached with the
bull of excommunication. The terrible Bemabo is supposed to have seized the bull,
perused it with darkening brow, then gazed for a moment down at the swift waters of
the river. Glaring at the envoys, he asked them if they would rather eat or drink. The
frightened ambassadors, knowing Bernabo's grotesque sense of humor and certain that if
they expressed a preference for drinking they would soon find themselves in the river,
chose to eat. They were accordingly seized by Visconti men-at-arms and forced to eat
the papal bull itself, parchment, seals and all, while their tormenter looked on with
sadistic amusement."

To this extraordinary story, which Cox included in the text of his book, the follow-
ing footnote was appended: "To paraphrase Winston Churchill, tiresome investigators
have undermined this excellent tale, but it should still find its place in any history worthy
of the name." Eugene L. Cox, The Green Count of Savoy (Princeton, 1967), p. 155.

FALLACIES OF FACTUAL SIGNIFICANCE 75

never what they seem to be, with a firm attachment to the doctrine of
original sin.

There is a little of the furtive fallacy in us all — enough to sustain
the common truth of Ralph Barton Perry's observation that "facts, like
sinners, gain something from an unsavory reputation." 26 But there is
more of it in some people than in others. And when there is much of
it, we are apt to summon a psychiatrist. In an extreme form, the fur-
tive fallacy is not merely an intellectual error but a mental illness
which is commonly called paranoia. 27

Sometimes paranoia is an epidemic disease. There are periods in
the past when men ran mad in packs, and when their madness took
precisely this form. Sometimes, such outbreaks have been followed by
great cataclysms. The relationship of cause to consequence is cloudy,
but a symptomatic significance, at least, seems clear. If it is correct,
then nothing is more ominous in our own time than the prevalence of
furtive fallacies in every ideological camp. Witness the appalling suc-
cess of Macbird, and the tone of the campaign against Axel Springer,
and miscellaneous fears of furtive acts by soldiers and students, Ne-
groes and Jews, Communists and capitalists, the Mafia, Maoists, the
Kennedy clan, President Johnson, General de Gaulle, and many others.
Reasonable men of all persuasions have an interest, and a duty, to
oppose this ugly tendency. The furtive fallacy has a self-fulfilling quality.
Men who believe it begin to act furtively. Nothing is more dangerous to
the peace of the world or more deleterious to the progress of humanity.

The furtive fallacy is, however, nothing new in the world. A
rounded history of this delusion would carry us back to the great Homeric
treacheries and far beyond. We might merely return sixty years, or so,
to muckraking journalists of America's "Progressive era." If Richard
Hofstadter is correct, they operated upon an assumption that reality
was a "rough and sordid" thing. "It was hidden, neglected, and off-stage.
. . . Reality was the bribe, the rebate, the bought franchise, the sale
of adulterated food. . . . Reality was a series of unspeakable plots, per-
sonal iniquities, moral failures." 28

Historians in the Progressive era showed the same habit of thought.
A socialist scholar, Algie Simons, led the pack in their reinterpretation
of that sacred ark of republicanism, the United States Constitution. In

26. Ralph Barton Perry, Puritanism and Democracy (New York, 1944), p. 53.

27. Of course, I do not mean to suggest that the historians whose works are used as
examples of this error are paranoid. The argument, rather, is that a fallacy which ap-
pears in their thought has sometimes become paranoia in the thought of others.

28. Richard Hofstadter, The Age of Reform, Vintage ed. (New York, 1960), pp. 201-2.

76

INQUIRY

rhetoric which "sweats with rural superstition," 29 he informed his
readers that "the organic law of this nation was formulated in secret
session by a body called into existence through a conspiratory [sic]
trick, and was forced upon a disfranchised people by means of dis-
honest apportionment in order that the interests of a small body of
wealthy rulers might be served." 30

Close behind Simons was the mightiest muckraker of them all,
Charles Beard, author of the most famous monograph in American
history, An Economic Interpretation of the Constitution (New York,
1913), in which the furtive fallacy was deeply embedded. Beard claimed
otherwise, and several times insisted that his thesis was misunderstood.
But in fact it was misconceived. For Beard, as for Simons, the Con-
stitution was "essentially an economic document," which was "written
by a small and active group of men" who were "with a few exceptions,
immediately, directly and personally interested in, and derived economic
advantages from, the establishment of the new system." 31

Beard's book, for all the controversy which it caused, was a very
moderate specimen of the furtive fallacy. No imputation of paranoia
can attach to his work. Nobody, not even his worst critics, wished to call
a psychiatrist (though President Nicholas Murray Butler of Columbia
may have wished to call the police). But his interpretative model, in
the events which it made significant, was false and misleading.

Beard's last major book, on the origins of American intervention
in World War II, was even more deeply flawed by the furtive fallacy
in its thesis that Franklin Roosevelt and his cronies secretly manipulated
American policy by a series of subtle and sordid tricks to bring their
nation into the war. There is no evidence that Beard deliberately
falsified his account. 32 The errors and distortions in his interpretation
are rather the result of his erroneous assumptions — not merely in the
conceptualization of the specific historical problem which he was study-
ing, nor simply in his political prejudices, but in the way he believed
history happened. That fundamental error was a constant in Beard's

29. Stanley Elkins and Eric McKitrick, The Founding Fathers: Young Men of the
Revolution (Washington, D.C., 1962), p. 8. I am heavily endebted to this essay, in the
following interpretation.

30. Algie M. Simons, Social Forces in American History (New York, 1911), p. 99;
quoted in Elkins and McKitrick, p. 8.

31. An Economic Interpretation of the Constitution, new ed. (New York, 1935), pp.
324-25; cf. pp. xvi, 73.

32. Cushing Strout notes that Basil Rauch in Roosevelt from Munich to Pearl Harbor
(New York, 1950) "attacks Beard's devil-theory of Roosevelt at the price of creating a
devil-theory of Beard by accusing him of deliberate falsification of the record." The
Pragmatic Revolt in American History: Carl Becker and Charles Beard (New Haven,
1958), p. 150 n.

FALLACIES OF FACTUAL SIGNIFICANCE

77

career, from the Economic Interpretation in 1913 to President Roosevelt
and the Coming of the War, 1941: A Study in Appearances and Reali-
ties (New Haven, 1948).

If the furtive fallacy is a link between the early Beard and the late
Beard, it is also a bond between Beard and the anti-Beardians. We have
already considered the fallacy of counterquestions in this connection.
The mistake perpetrated by Beard was perpetuated by his most able
critic. In Forrest McDonald's history of the Constitution, there is a furtive
fallacy which is so gross that I have sometimes wondered if it was in-
tended as a practical joke.

McDonald tells us, with a straight face, that the critical compromise
which made the Constitution possible was a secret deal between Con-
necticut land speculators, represented by Roger Sherman, and South
Carolina planters, represented by John Rutledge. The arrangement was
made in a smoke-filled room, on the evening of June 30, 1787, but al-
most came unstuck in another smoke-filled room, where Charles Pinck-
ney (who was in on the secret), and Luther Martin (who was not),
went off on a great roaring drunk together. This alcoholic bout might be
called the first anti-Federal party. Sometime after the seventh drink,
bibulous "blackguard-Charlie" whispered the secret to the cunning
"Brandy Bottle" Martin, who proceeded to rally the opposition. But
Oliver Ellsworth and John Rutledge, who are compared to Samuel Ray-
burn and Lyndon Johnson, managed to put the deal back together again,
by some sly parliamentary skulduggery. 33

This remarkable interpretation is developed with imaginative detail,
in a work which might be enjoyed as a masterpiece of black humor,
worthy of comparison with the best of Terry Southern (with James
Madison in the role of Candy), or Pynchon (with the Constitution as
V), or Donleavy (with Robert Morris as Sebastian Dangerfield), or
John Barth himself. I suppose that it might have happened — even Barth's
scatalogical secret history of Captain John Smith might have happened.
But there is no evidence to sustain McDonald's interpretation, and much
to the contrary. There is no good evidence even that the alleged meetings
took place, much less that such bargains were made.

This furtive fallacy is central to McDonald's interpretation. But
there are also many smaller ones, in a work which is rum-and-strumpet
history, with a vengeance. What, for example, really went on, in that
green paradise called New Hampshire? McDonald believes that reality
was something like this: "Winters in eighteenth-century New Hampshire
were uncommonly long and cold: so cold that all men save the most

33. Forrest McDonald, E Pluribus Unum: The Formation of the American Republic,
1776-1790 (Boston, 1965), chap. 6, pp. 176-84.

78 INQUIRY

industrious stayed indoors, making their wives pregnant and praying to
a Fundamentalist, Calvinist God to make it warmer; and so long that
when Spring finally broke, all men save the most industrious got fiercely
drunk and made other men's wives pregnant" (p. 114). Who really was
an obscure Maryland politician, Daniel of St. Thomas Jenifer? "Jenifer
was a petty local official who customarily had his hand in the public till
or was helping someone else's to reach it" (p. 166). What was the real
story behind the Jay-Gardoqui treaty? "Gardoqui flattered Jay and show-
ered Jay's wife (whom Jay adored) with attention and gifts, whereupon
Jay agreed to ask Congress for permission to surrender American claims
to rights of navigation on the Mississippi" (p. 82).

In the misbehavioral science of Beard and McDonald, reality is
always the underlying fact, always something more than meets the eye.
But there is also another curious inversion of the furtive fallacy, which is
grounded in an increasingly fashionable assumption that reality is some-
thing less than meets the eye. An example is a work by Joseph Hamburger
on the philosophic radicals and the English reform bill. His thesis, as
summarized by a critic, is that "their threat of violence and revolution
was a calculated bluff, a threat they neither intended nor desired to carry
out, and that their strategy succeeded not because everyone was taken in
by the bluff, but because some politicians, recognizing it as such, chose
to submit to it for tactical reasons of their own." Reality is reduced to a
set of shadows, flickering behind a curtain of flimsy rhetoric. 34

£*> The moralistic fallacy selects edifying facts. In Michael
Oakeshott's phrase, it makes the past into "a field in which we exercise
our moral and political opinions, like whippets in a meadow on Sunday
afternoon." 35

In the fields and meadows which surround most American uni-
versities, there are hordes of hairy graduate students who are crying up
this error, as a Whole New Thing in historical scholarship. But it is a
very old thing — as ancient as Dionysius of Halicarnasus, who character-
ized history as "philosophy by examples." And it is inconsistent with a
serious and disciplined empirical inquiry into what actually happened. It
would make history a handmaid of moral philosophy, as it actually was
in most American colleges, during much of the nineteenth century. 36

34. Gertrude Himmelfarb, American Historical Review 61 (1966): 1344. The book is
Joseph Hamburger's James Mill and the Art of Revolution (New Haven, 1963). See
pp. 115, 125.

35. Michael Oakeshott, Rationalism in Politics and Other Essays (London, 1962),
p. 165.

36. See, e.g., Wilson Smith, Professors and Public Ethics (Ithaca, N.Y., 1956).

FALLACIES OF FACTUAL SIGNIFICANCE

79

This fallacy is exceedingly difficult to define precisely. Most reason-
able men would agree that all historians, without exception, must and
should make value judgments in their work. But they would also agree
that some historians moralize upon past events in ways which are in-
consistent with empiricism. There is a thin line, somewhere in between,
which is not easy to locate.

Historians with whom I have discussed this problem have variously
suggested five separate criteria for distinguishing acceptable moral judg-
ments from unacceptable moralizing:

1 . Moral judgments are good, true, or sensitive judgments; moraliz-
ing is bad, crude, shallow, false, or stupid.

2. Moral judgments are moderate; moralizing is immoderate.

3. Moral judgments are empirically accurate; moralizing is em-
pirically false.

4. Moral judgments consist in judging people by their own stand-
ards; moralizing is judging them by other people's standards.

5. Acceptable moral judgments are functional, or neutral, to
empirical inquiry; unacceptable moralizing is dysfunctional.

All but one of these notions seem clearly untenable. The first, I think,
is circular. Though it conforms to common usage among historians, it
tends to become, in practice, mere special pleading, like the statement
"I make moral judgments; you moralize; he is a self-righteous prig." The
second distinction is also invalid, because it makes moderation the mea-
sure of truth. It is a classic example of argument ad modum, discussed
in chapter XI below.

The third distinction seems equally unacceptable in its premise that
ethical problems are susceptible to empirical resolution. Without plung-
ing into the philosophical thickets of ethical objectivism, which I am not
qualified to discuss, it appears to me that moral judgments are, in context,
a priori in their nature. The fourth distinction implies an ethical rela-
tivism which is as hateful and suicidal as historical relativism.

By process of elimination, we are left with the fifth distinction, which
makes more sense to me than any other. What does one mean by moral
judgments which are functional or neutral or dysfunctional to empirical
inquiry? Merely this: every historian possesses a complex structure of
value assumptions, which he cannot adjust to his empirical projects, and
cannot keep out of his work. But he can adjust his project to his values
in such a fashion as to neutralize or to control his moral preferences. The
first step in that process would be to make his values as fully explicit, to
himself and others, as possible. The second step would be to design a
research problem in which his values allow an open end.

80 INQUIRY

Imagine two historians, one a southern white supremacist and the
other a black militant. Each possesses many values which are separate
from white supremacy or black militancy, but we will confine our at-
tention to these elements alone. (The nature of the problem does not
change as it becomes more complex. ) Neither of these scholars, however
honest and well-intentioned they may be, can ever hope to write a
rounded general history of Negroes in Reconstruction without moralizing
upon their subject in a dysfunctional way.

Our hypothetical black militant wishes, of course, to write a history
of Blacks. Within the general area, he can find problems in which his
moralizing propensity is controlled. He might try his hand, for example,
at a comparative study of black slavery and anti-black prejudice in
Spanish, French, Dutch, Portuguese, Danish, and English colonies — a
study which we very much need at the present moment. His values would
be engaged in the problem, but functionally engaged, for he has an open
end in his inquiry. Our imaginary southern white supremacist, on the
other hand, wants to study the history of the American South. Maybe he
could do a comparative analysis of regional white culture patterns within
the South and bring a functional intensity of moral purpose to his work
without moralizing upon it in a dysfunctional way.

We might, in light of these observations, refine our first, approxi-
mate, definition of the moralistic fallacy, which becomes not merely, in
Michael Oakeshott's metaphor, exercising our moral opinions in history
like a whippet in a meadow on Sunday afternoon. It is exercising the
whippet without a leash.

A classical example of an unleashed moralizer is a great Roman
historian, Tacitus, who possessed a glorious gift for vivid and, I think,
exact analysis of historical problems. His remarkable characterizations,
like the Roman sculpture of the same period, gain a universal meaning
from their profound particularity. But that splendid talent coexisted with
an incompatible purpose. "This I regard as history's highest function,"
he wrote, "to let no worthy action be uncommemorated, and to hold out
the reprobation of posterity as a terror to evil words and deeds." 37 Tacitus
gave this moralizing method a famous twist in the Germania, where the
chastity, simplicity, freedom, and fidelity of a primitive people are set up
as an antithesis to the debauchery and corruption which he despised in
his own culture. The book is punctuated with sarcastic side comments
about the ignoblest Romans of them all. Classical historians have argued
endlessly about the priority of his purposes. Did he mean to describe
Germans as they were, or Romans as he wished them to be? The only

37. Tacitus, Annals, 3. 65.

FALLACIES OF FACTUAL SIGNIFICANCE

81

sensible answer, I think, is that he tried to do both at the same time, and
therein lies the fallacy, for his moral judgments were dysfunctional to
his historical interpretation of the Germans. 38

From the age of Tacitus to our own time, the moralistic fallacy
has flourished in a thousand forms. My favorite is The Mirror for Magis-
trates, a versified raid upon English history, probably first conceived in
the twelfth century, and much enlarged by "dyuers learned men" through
the next four hundred years. The work was addressed to the "nobilitye
and all other in office," in the hope that "For here as in a loking glas,
you shall see (if any vice be in you) howe the like hath bene punished
in other[s] heretofore, whereby admonished, I trust it will be a good
occasion to move you to the soner amendment. This is the chiefest ende,
whye it is set furth, which God graunt it may attayne." 3 '

In the moral universe of the Mirror, everyone received his just
reward:

Humber the king of the Hunnes shewes how he minding to conquere this
land was drowned, &c. . . . Morindus a bastarde, declares how hee was
exalted to the kingdome, waxed cruell, and at laste was deuoured by a
monster, the year before Christe 303. . . . King Emerianus for his tirany
was depose, about the year before Christ, 225. . . . King Chirinnus giuen
to dronkennesse raygned but one year. Hee died about the yeare before
Christ, 137. 40

A noble gentleman who glanced into the Mirror could see "How
Guiderious . . . became desirous to winne all the worlde, spoyled France,
Germany and a great part of Italy: and lastly, how hee was miserably
slayne in a tempest of thunder. . . . This history is a synguler ensample of
God's vengeance, against pride and arrogancy." He could learn "Howe
Vter Pendragon was inamoured of Duke Garelus wife, and howe by
laweless loue he lost his kingdome. This example is most necessary for
the present time. ..." and "How Sigebert was thrust from his throne,
and miserably slayne by a Heardman. This tragedie dooth teach both
Prince and suiect his duetie at large." 41

The "ensampling" method of the Mirror for Magistrates, of course,
is as faithful a representation of the reality of history as a fun-house
mirror. For every sinner who has been struck by lightning, or provi-
dentially slain by a shepherd, a hundred have lived happily ever after.

38. A juxtaposition of primitive virtue and civilized corruption is a common variant
of the moralistic fallacy, which mars many modern works in history and anthropology.
It is deeply embedded in the curious cult of "structuralism" which has formed in France
around the work of a great ethnologist, Claude Levi-Strauss.

39. Lily B. Campbell, ed., The Mirror for Magistrates (Cambridge, 1938), pp. 65-66.

40. Lily B. Campbell, ed., Parts Added to The Mirror for Magistrates (Cambridge,
1946), pp. 74, 185, 282, 283.

41. Ibid., pp. 386, 452.

82 INQUIRY

But, alas, even that fact has been made into a moralistic fallacy by the
Swiss historian, Jacob Burkhardt. "Every violent deed that prospers is,
to say the least, an offense; that is to say, a bad example. The only thing
to be learned from successful misdeeds of the mighty is not to value life
in this world more highly than it deserves." 42

The pragmatic fallacy selects useful facts — immediately and
directly useful facts — in the service of a social cause. Most historians
hope that their work is, or will be, useful to somebody, somewhere,
someday. In the conclusion we shall consider how utility can function in
history. But the pragmatic fallacy short-circuits the problem. It consists
in the attempt to combine scholarly monographs and social manifestoes
in a single operation. The result is double trouble: distorted monographs
and dull manifestoes.

Consider a recent work by Eric Williams, a trained scholar and a
prominent West Indian statesman, who explicitly pursued two entirely
different and incompatible purposes. The first was to do an "accurate,"
"adequate," and "informed" history, in which the author attempted, in
a Rankean spirit, "to let the documents speak for themselves." 43 But the
second was something else: 'The aim in writing the book, however, was
not literary perfection or conformity with scholarly canons," Mr. Williams
wrote. "The aim was to provide the people of Trinidad and Tobago on
their Independence Day with a National History, as they have already
been provided with a National Anthem, a National Coat of Arms, Na-
tional Birds, a National Flower and a National Flag." 44

These emblems, for Williams, were meant to be both useful and
ornamental. He industriously collected facts which promoted two parti-
cular purposes: the "integration of the races" in Trinidad and Tobago
and "integration of the separated Caribbean territories." The "national
history" of Trinidad and Tobago which resulted, closely resembled its
national flag in more respects than one — a great heavy black diagonal,
snug between two thin bands of white ( Spaniards and British, I presume) ,
all superimposed upon a red field (Williams's interpretation is predomi-
nantly Marxist). Facts are selected to show that "in Trinidad the Negro,
the Indian, French and Spaniard, English and Portuguese, Syrian and
Lebanese, Chinese and Jew, have all messed out of the same pot, all
are victims of the same subordination, all have been tarred with the

42. Quoted by Ernst Cassirer in The Problem of Knowledge, p. 275.

43. Eric Williams, History of the People of Trinidad and Tobago (Port-of -Spain, 1962),
p. vii.

44. Ibid.

FALLACIES OF FACTUAL SIGNIFICANCE

83

same brush of political inferiority.. Divergent customs and antipathetic
attitudes have all been submerged in the common subordinate status of
colonialism." 45 And that thesis is as false and misleading as it is possible
for a thesis to be.

But it is not more false or misleading than other patriot histories
which have been written in all nations for the same purpose of national
integration. There is no appreciable difference in this respect between
Williams' interpretation and that of national historians of Britain, who
equally condemned Oliver Cromwell as a great bad man and James II
as a tyrant-fool, all in the cause of nation-building; or American histo-
rians — maybe the worst of the lot — who have filled the heads of little
children with functional fairy tales about the good Pilgrim fathers and
wicked witch-buming Puritans; about bad King George and the demigods
who wrote the Constitution, the Monroe Doctrine, the Great Emancipa-
tor, the Glorious Melting Pot, and all the dirty Indians, Frenchmen,
English, Scots, Hessians, Algerians, Mexicans, Spaniards, Filipinos,
Nicaraguans, Prussians, Italians, Japs, Koreans, Chinese, and Vietnam-
ese who had to bite the dust so that the world would be made safe for
Peace, Brotherhood, and the Universal Yankee Nation. The fairy tales
I was taught, in a schoolroom below the Mason and Dixon line, were a
little different. But the fallacy was much the same.

Today, there isn't much of this nation-building variant of the prag-
matic fallacy in the serious scholarly literature of the United States and
Britain. But the same form of error is still very common in the service of
other causes. A case in point is the work of a young radical American his-
torian, Staughton Lynd. He is a scholar who is possessed of extraordinary
gifts — intelligence, industry, imagination, integrity, candor, a sense of per-
spective, a sense of proportion, and even a sense of humor. By temperament
and conviction he is not an observer but a participant. He enthusiastically
endorses an aphorism paraphrased from Marx's Theses on Feuerbach:
"The Historians have interpreted the world; the thing, however, is to
change it." 46 To this declaration, Lynd adds a rhetorical question of his
own: "Should we be content with measuring the dimensions of our prison
instead of chipping away, however inadequately, against the bars?" 47

For many years Lynd has been working like a beaver at the prison
bars — not with a file but with a file card. 48 It would appear to be arduous

45. Ibid., p. 280.

46. Staughton Lynd, "Historical Past and Existential Present," in Theodore Roszak, ed.,
The Dissenting Academy (New York, 1967), p. 107.

47. Ibid.

48. The task is tough enough, given Lynd's choice of tools. But the trouble is com-
pounded by the fact that Lynd and his radical friends believe that the great penitentiary
in which they live — i.e., American society — is cunningly constructed with invisible bars.

84

INQUIRY

and unrewarding toil. But Lynd is determined to demonstrate that a
"professor of history" (in a literal sense) should be a historical protago-
nist in one and the same act. He believes specifically that a historian should
promote the transcendent purpose of class revolution in the American
republic. Now, to have a class revolution, one must have class conflict.
One major group of his writings is organized around this object. For
Lynd, a significant fact is a useful fact, and a useful fact is a fact which
demonstrates the reality of class conflict in America and even promotes
it. To this end he has published a monograph and several articles, mostly
about affairs in Dutchess County, New York, mainly in the period 1 760-
1790. It is work of high professional quality which genuinely deals in
facts — true statements about the past. But there is a fallacy in the true
statements which Lynd chooses to make and in the criteria of factual
significance upon which he bases his choice. The author himself is his
own best critic. Writing five years after the monograph was published,
he candidly observes,

I am now more conscious that I selected a range of data which I could
be pretty certain would substantiate the thesis I hoped was true. I studied
opposition to the United States Constitution in Dutchess County, New
York, because Dutchess County had a history of landlord-tenant conflict
very likely to be connected with how groups aligned themselves for or
against ratification of the Constitution. The bias involved in my selection
of Dutchess County did not necessarily invalidate my findings, but it
raised serious question as to their generalizability. I believe this is how
bias characteristically operates in the work of other historians, too: not
in the deliberate mishandling of evidence, but in the selection of research
design. 40

Other batches of Lynd's essays are devoted to the discovery of
other kinds of useful facts. One group of them represents an attempt
to locate an authentic radical tradition in American history. "As one
considerably alienated from America's present," he writes, "I wanted to
know if there were men in the American past in whom I could believe." 80
A third group of articles, which are functionally related to Lynd's

He has worked for years at what he thinks is an invisible bar, only to discover that it
is really a space between them, and that the work must begin again. This cycle has
been repeated several times in Lynd's scholarly career.

49. Ibid., p. 10. The monograph is Anti-Federalism in Dutchess County, New York:
A Study of Democracy and Class Conflict in the Revolutionary Era (Chicago, 1962).
The articles are reprinted in Class Conflict, Slavery, and the United States Constitution
(Indianapolis, 1967).

50. "A Profession of History," p. 10. These works include Nonviolence in America:
A Documentary History (Indianapolis, 1966); and "Beard, Jefferson and the Tree of
Liberty," in Class Conflict, Slavery, and the United States Constitution, pp. 247-69; and
Intellectual Origins of American Radicalism (New York, 1968).

FALLACIES OF FACTUAL SIGNIFICANCE

85

participation in the civil rights movement, centers upon Negro slavery in
American history. Lynd argues that the Peculiar Institution was a key,
maybe the key to American history, the significance of which has been
underestimated by American historians. He suggests that the issue of
slavery was the issue in the politics of the early republic, and particularly
in the framing of the Constitution. In these articles, I think, unlike the
Dutchess County essays, Lynd does not make true statements. He means
to make them, but his pragmatic criteria of factual significance lead him
to exaggerate grossly a few scraps of evidence, mostly circumstantial,
into an interpretation, mostly false. Most American historians would
surely agree that slavery was the predominant issue in American politics
from 1820 to 1860. But to argue that it was equally so in the preceding
forty-year period is simply wrong. 51

An introduction to one of Lynd's collections of essays has been
written by a radical English historian, whose works provide many other
examples of the same sort of error. E. P. Thompson is a little more
heavy-handed than his American colleague. His introduction is a ritual
incantation against "mutton-fisted narks of academe," but he is also a
serious and gifted scholar who has written a work of great importance
on the English working class. 52 His work, however, is seriously flawed
by his standards of factual significance, which lead him to select facts
which are useful to the preservation of the culture and identity of the
working class against a variety of enemies, imagined and real. 63

Radical historians have no monopoly on the pragmatic fallacy. The
same error is committed by historians who are located at many dif-
ferent points in the spectrum of social ideology. A familiar and flagrant
example is the recent work of Arthur Schlesinger, Jr., in his writings on
the New Deal. 54 Turning to a conservative historian, we find a different
form of the pragmatic fallacy, as well as a different substance — a form
which is not merely deleterious in its methodological consequences but
exceedingly dangerous as well. Scholars who take a pragmatic view of
their task and collect facts that are weapons for a cause are faced with
the problem that some facts exist which are useful to their enemies. More
than a few able historians, caught up in this predicament, have proposed
a kind of fact control, which is profoundly hostile to free inquiry.

In the early stages of the Cold War, the American Historical As-

51. Lynd, "On Turner, Beard and Slavery," and "The Compromise of 1787," in Class
Conflict, Slavery, and the United States Constitution, pp. 135-52, 185-213.

52. The Making of the English Working Class (New York, 1964).

53. Another excellent work, similarly flawed, in much the same way, is Richard
Hoggart, The Uses of Literacy (London, 1957), p. 280.

54. See, e.g., The Coming of the New Deal, pp. 1-2, 16, 20-23, 102, 169-70, 175-76,
passim.

86

INQUIRY

sociation elected as its president Conyers Read, a distinguished historian
of Tudor England, a deep believer in "the relativity of all history," and
a dedicated anti-Communist. Seriously concerned about the fate of free
institutions in the postwar world, he advocated a little fact control in
historical scholarship.

We must assert our own objectives, define our own ideals, establish our
own standards and organize all the forces of our society in support of
them. Discipline is the essential prerequisite of every effective army whether
it march under the Stars and Stripes or under the Hammer and Sickle.
. . . Total war, whether it be hot or cold, enlists everyone and calls upon
everyone to assume his part. The historian is no freer from this obligation
than the physicist. ... If historians, in their examination of the past,
represent the evolution of civilization as haphazard, without direction and
without progress, offering no assurance that mankind's present position is
on the highway and not on some dead end, then mankind will seek for
assurance in a more positive alternative, whether it be offered from Rome
or from Moscow. . . . This sounds like the advocacy of one form of social
control as against another. In short, it is. But I see no alternative in a
divided world. Probably in any planned world we can never be altogether
free agents, even with our tongue and our pen. The important thing is
that we shall accept and endorse such controls as are essential for the
preservation of our way of life. . . . This need not imply any deliberate
distortion of the past in the interests of any ideology. Always it will be
our obligation as historians to consider the present developing civilization
in all its aspects. We shall still, like the doctor, have to examine social
pathology if only to diagnose the nature of disease. But we must realize
that not everything which takes place in the laboratory is appropriate for
broadcasting at the street corners. ... I am inclined to think that the first
prerequisite of a historian is a sound social philosophy. Actually he finds
in the past what he looks for. 55

Professor Read, now in his grave, was a scholar and a gentleman.
But his doctrine is utterly despicable and very dangerous to historical
science. It is, I fear, one of the many poisoned fruits of relativism and
of the Cold War altogether. And it is still very much alive.

There is something equally ugly in American radicalism, of which
the principal manifestation is a book by Herbert Marcuse and others,
called A Critique of Pure Tolerance (Boston, 1965), in which Mr.
Marcuse demands "new and rigid restrictions on teachings and practices
in educational institutions," and specifically the withdrawal of tolerance
from anyone who promotes "armament," "chauvinism," or "discrimina-
tion on the grounds of race and religion," or who opposes the "extension
of public services, social security, medical care, etc."

The spectacle of an impotent radical minority threatening to with-

55. Conyers Read, "The Social Responsibilities of the Historian," American Historical
Review 55 (1950): 283-85.

FALLACIES OF FACTUAL SIGNIFICANCE

87

draw tolerance from the moderate mass would be amusing if it were not
so profoundly hostile to a principle which has been very slow to take root
in the world, and which is still very fragile. Experience suggests that the
repressive form of the pragmatic fallacy is highly contagious — it can
spread swiftly if it is not actively contained.

The aesthetic fallacy selects beautiful facts, or facts that can
be built into a beautiful story, rather than facts that are functional to
the empirical problem at hand. It consists in an attempt to organize an
empirical inquiry upon aesthetic criteria of significance, or conversely in
an attempt to create an objet d'art by an empirical method. To do so is
to confuse two different kinds of knowledge and truth.

To the truth of art, external reality is irrelevant. Art creates its own
reality, within which truth and the perfection of beauty is the infinite
refinement of itself. History is very different. It is an empirical search for
external truths, and for the best, most complete, and most profound
external truths, in a maximal corresponding relationship with the absolute
reality of the past events. Any attempt to conduct that search according
to aesthetic standards of significance (most commonly in an attempt to
tell a beautiful story) is either to abandon empiricism or to contradict
it.

The aesthetic fallacy is an ancient form of error, which appeared
full blown in Aristotle's notorious opinion that history was an inferior
form of particularized poetry, roughly on an intellectual plane with lyre
plucking. 56 That Aristotle should have come to this conclusion is under-
standable, for the aesthetic fallacy appears in the writing of every ancient
historian whose works have survived into the modern age, particularly
in historians of the fourth century B.C. 57

But the logical extension of the aesthetic fallacy appears later, in the
Roman historian who is reputed to have said that he would have made
Pompey win the battle of Pharsalia if the turn of the sentence had re-
quired it. This extraordinary statement is attributed to Livy by Lytton
Strachey, who thought it a perfectly sensible idea. 68

56. Poetics, ed. Richard McKeon, The Basic Works of Aristotle (New York, 1941),
chaps. 1, 4, 9.

57. J. B. Bury, The Ancient Greek Historians (New York, 1958), pp. 160-78. An ex-
ception is a great fifth-century historian, Thucydides, who specifically condemned the
errors of poets "who exaggerate the importance of their themes," and wrote his speeches,
especially, in a style which is a schoolboy's despair, a Greek style so crabbed and ugly
that one wonders if he contrived it deliberately, to demonstrate that he wished to tell
the unvarnished truth.

58. Lytton Strachey, Spectatorial Essays (London, 1964), p. 13.

88 INQUIRY

The attribution is probably incorrect. Indeed, one wonders if Livy
was made to say it for the turn of Lytton Strachey's sentence. But, there
was, in truth, more than a little of this madness in Livy's method. A learned
student of his works has compared him to Sir Walter Scott. "Like a
novelist," writes R. M. Ogilvie, "he subordinated historical precision to
the demands of character and plot. He indulged freely in invention and
imagination in order to present a living picture. He would have dis-
claimed the title of a 'Historian' in a modern sense." 69

Strachey's alleged quotation would be even more accurately
descriptive of the Roman poet-historian Lucan than Livy. Lucan actually
wrote a long epic poem, commonly called the Pharsalia. Though he did not
reverse the outcome of the battle for poetic effect, he revised nearly
everything else, in the interest of aesthetic perfection. For the sake of his
meter he rearranged the geography of the Mediterranean, substituting
Emathios for Thessalios in the first line, because Thessalios would not
scan after the opening Bella per.™ He removed the capital of Parthia
from crude, clumsy-sounding Ctesiphon, to beautiful, euphonious Baby-
lon. 61 Like many another Roman poet after Virgil, he made the battles
of Pharsalia and Philippi take place on the same spot, though more than
150 miles lay in between. He changed the sequence of consuls when it
suited his poetic purpose, promiscuously mixed real and fictitious per-
sons, improved the inconvenient topography of Greece, reversed the
Persian Gulf and the Red Sea, revised the career of Pompey, and brought
Cicero to the battlefield for dramatic effect. 62

Lucan is long gone and little mourned. But the aesthetic fallacy
appears to be immortal. There are many examples in the modern world
of both historians and novelists who have ignored Virginia Woolfs sound
maxim that "truth of fact and truth of fiction are incompatible." 63 Two
able and deservedly distinguished modern historians may serve as ex-
amples: first, A. L. Rowse, an English historian who rarely publishes a
volume without a prefatory barrage of angry epithets, aimed at academic
colleagues who fail in his judgment to fulfill the aesthetic potential of
their discipline. Nobody can read, or rather study, a representative doc-
toral thesis without sympathizing in some degree. But Rowse seems to
have a curious conception of his discipline. "History," he writes, "is a
great deal closer to poetry than is generally realised; in truth, I think, it
is in essence [that word!] the same." 64 Rowse's meaning, in context, is

59. The Listener, Nov. 3, 1960.

60. Robert Graves, ed., Lucan, Pharsalia (Baltimore, 1957), p. 17.

61. J. D. Duff, ed., Lucan (London, 1951), p. 3.

62. Duff, ed., Lucan, pp. 52-53, 268, 320, 328, 359, 369, 372, 412.

63. Virginia Woolf, Collected Essays, 4, 234.

64. A. L. Rowse, The Use of History (London, 1946), p. 55.

FALLACIES OF FACTUAL SIGNIFICANCE

89

unclear. But I understand him to argue the ancient Aristotelian proposi-
tion that history is particularized poetry and to urge upon his colleagues
the double standard of beauty and empirical truth — a double mode of
selection, in which events are significant both for their functional rela-
tion to the historical problem, and for their aesthetic quality. But the two
standards are incompatible, and some of Rowse's work falls between
them. A reviewer of one of his books complains that "details are often
murky and sometimes inaccurate; biased judgments and dubious points
of history are frequently put forward as established fact." 05

A second example is Bernard De Voto, an American novelist who
became an American historian. But the conversion was incomplete. He
called history a "spectacle" — something to be staged for the aesthetic
gratification of the reader. And he undertook to write a book about
American events in the year 1846, "because 1846 best dramatizes
personal experience as national experience." 96 The book characteristically
begins with a splendid story, which De Voto describes as "doubtless more
beautiful than true." It concerns a group of American trappers in the
spring of 1846, who, after a violent thunderstorm, saw the image of an
eagle in the red western sky, superimposed upon the sun. This tale is
followed by another baroque paragraph describing an omen which is
equally spectacular in its literary effect: Biela's comet, in that annus
mirabilis, split in two on its way toward perihelion, thus prefiguring the
fate of the Republic.

Throughout the rest of the book, there is much more in the same
spirit. The great Western eagle screams and flaps his wings, and lots
of brave but wretched little Mexicans scurry to and fro across the coun-
tryside. History becomes alternately a tragic "drama" of heroes who
are undone by their virtues, and a low "vaudeville show of swollen ego-
tism, vanity, treachery, incompetence, rhetoric, stupidity and electioneer-
ing." De Voto clearly makes every effort to be accurate. But one cannot
read his books without concluding that he is torn between two incon-
sistent objects — to tell what actually happened, and to tell a beautiful
story. In pursuit of his two purposes, he operates at cross-purposes and
diminishes the degree to which either is fulfilled. 87

A third example is also an American historian, Samuel Eliot Mori-
son, who believes deeply in "history as a literary art." In an essay pub-

65. Darrett B. Rutman, in William and Mary Quarterly, 3d ser. 18 (1961): 134-36;
reviewing The Elizabethans and America (New York, 1959).

66. Catharine Drinker Bowen, et al., Four Portraits and One Subject (Boston, 1963),
pp. 1-27.

67. Bernard De Voto, The Year of Decision (Boston, 1943), pp. 4-5, 234, 496-97
passim.

90 INQUIRY

lished under that title, 68 he urges young historians not to "bore and
confuse the reader with numerous 'buts,' 'excepts,' 'perhapses,' 'howevers,'
and 'possiblies.' " Instead they are advised to tell beautiful stories for the
aesthetic intoxication of their readers. But this, I would think, would be
storytelling in more senses than one. A historian who omits all the ugly
"buts," "excepts," "perhapses," etc., is falsifying the record. I do not
mean to suggest that Morison is a deliberate falsifier — far from it. But his
method has a falsifying effect. It is profoundly dysfunctional to serious
historical study.

The quantitative fallacy is the latest form of insignificance,
which consists in the idea that the facts which count best count most.
It should not be confused with quantification proper — an important
tool, long used by historians and presently in process of a revolutionary
refinement. 69 To "quantify" is merely to count; "quantifiers" thus include
schoolboys who are taught to tally the Five Points of Calvinism on their
fingertips, as well as scholars who employ sophisticated techniques of
regression analysis, and an IBM computer. Surely no reasonable man will
deny that counting has always been useful to historical inquiry and that
it will ever continue to be so; and that every historian should count
everything he can, by the best available statistical method.

But the quantitative fallacy is something else: a criterion of sig-
nificance which assumes that facts are important in proportion to their
susceptibility to quantification. There is an epigram, perhaps apocryphal,
attributed to Lord Kelvin, that everything which exists, exists in quantity.
Enthusiastic quantifiers have amended Lord Kelvin's statement to read,
"Unless a thing can be measured quantitatively, it does not exist sig-
nificantly." Therein lies a fallacy.

There are many significant things in the world today that nobody
knows how to measure. Someday, maybe, somebody will. But in the
meantime one must acknowledge their existence. Many ideational and
emotional problems, which lie at the heart of historical problems, cannot
be understood in quantitative terms. To move to the periphery, because

68. Samuel Eliot Morison, "History as a Literary Art," By Land and By Sea (New York,
1953) pp. 289-298, 294.

69. For an early specimen of quantification, see A. Lawrence Lowell, "The Influence of
Party Upon Legislation in America," American Historical Association, Annual Report
for 1901, 319-542, a roll-call analysis of party regularity in the British Parliament,
1836-1899, and the U.S. Congress, 1844-1899. In professional discussions of quantifica-
tion, as in public discussions of automation, there has been a good deal of nonsense
about the newness of it all, and an appalling lack of historical perspective.

FALLACIES OF FACTUAL SIGNIFICANCE

91

things can be measured there, is to behave like the man in Abraham
Kaplan's parable.

"There is a story," Kaplan writes, "of a drunkard searching under
a street lamp for his house key, which he had dropped some distance
away. Asked why he didn't look where he had dropped it, he replied,
'It's lighter here!' " 70

To make this argument is not to endorse the usual humanistic
blather about quantification as the enemy of the human spirit. There are
many critics of quantification in history who hold with Carlyle that "He
who reads the inscrutable book of Nature as if it were a Merchants
Ledger, is justly suspected of having never seen that Book, but only
some school Synopsis thereof; from which, if taken for the real book,
more error than insight is to be derived." 71 The argument here is a differ-
ent one — that the Book of Nature, like a merchant's ledger, might be kept
in double-entry style, one column listing phenomena which can be
quantified and another listing things which can be qualitatively known.
The latter is always longer than the former and can never be dismissed.

Criteria of significance should not be methodological, but substantive
in nature. They should always be grounded in the nature of the problem
itself and not in the tools of problem solving. The purpose of historical
inquiry is not to vindicate a method but to discover what actually hap-
pened. Every efficient means to this end is legitimate, but none alone
can be erected into a standard of legitimacy.

Behaviorists have reacted violently to this argument without meet-
ing it.

For some time [writes Warren E. Miller] it has been fashionable to argue
against the behavioral mode of research by insisting that one must beware
of becoming a prisoner of one's methods. The argument comes most
often from the traditional scholar who notes correctly that behavioral
science tends to ignore questions that accumulated wisdom has properly
defined as important. The argument is seldom sensitive to the possibility
that traditional workways often have taken a nettle for a rose and have
ignored many other questions — or have answered them badly — because
the methods were not equal to the task of mastering available information. 72

But this is an ignoratio, compounded by an ad hominem. The complaint
is not against the behavioral method of research, but rather against mak-

70. Abraham Kaplan, The Conduct of Inquiry (San Francisco, 1964), p. 11.

71. Thomas Carlyle, "On History," Critical and Miscellaneous Essays, 5 vols. (New
York, 1899), 2:91.

72. Warren E. Miller, "Promises and Problems in the Use of Computers: The Case of
Research in Political History," in E. Bowles, ed., Computers in Humanistic Research
(Englewood Cliffs, N.J., 1967), p. 84.

92 INQUIRY

ing that method into a standard of substantive significance. And Miller
cannot meet it by entering a complaint against somebody else's method.

In the past few years, historians who have issued manifestoes for
quantification have, by and large, been remarkably moderate in their
claims for that method. The restraint of W. O. Aydelotte, Samuel Hays,
and even Lee Benson, and the generally sympathetic, if not entirely
enthusiastic, response of their colleagues, is a hopeful sign that history may
avoid that unfortunate polarization into behavioralism and antibehavior-
alism, which largely explains the appalling poverty of political science as
a serious discipline during the past decade.

But scholars who have used quantification in historical research have
sometimes tended to be a little less cautious. One example is Jackson
Main's The Social Structure of Revolutionary America (Princeton,
196S), an important and useful inquiry into social class in America,
circa 1760-1790. Main's work has none of the most refined and esoteric
symbols which are the hallmark of a card-carrying quantifier. Never-
theless, his book is quantitative in its method and a specimen of the
quantitative fallacy. Main largely confined his attention to things he
could count — income and property, consumption patterns, occupations,
economic mobility. Even in a chapter on "contemporary views of class,"
he tended to count the number of classes which Americans recognized. In
consequence of his tacit assumption that things which count best count
most, his conclusions were both inaccurate and a little superficial. Main
believes, with many other scholars, that "the indigenous class structure"
of Revolutionary America "was based upon property rather than inherited
status." 73 That statement is surely true in some degree — but in what
degree? Property is something which can be counted with comparative
ease. Inherited status, however, involves such problems as deference,
which are much more difficult to manage in quantitative terms. Main's
quantitative bias, in his criteria of significance, tends to skew his conclu-
sions toward a property-based class system and away from the component
of hereditary status. And in so far as he dealt with hereditary status, he
tended to do so in terms of material emblems which inaccurately repre-
sent the problem. Main would have had to shift to a qualitative strategy
to attain a more fair and accurate approximation of his subject.

The superficiality of Main's book is equally serious and also closely
related to his quantitative bias. A reviewer writes:

That the book lacks a greater importance is in part the result of an exagger-
ated insistence on quantification which weakens rather than strengthens
one's confidence in the conclusions reached and limits the subtlety of
both the questions asked and the answers given. On certain occasions,

73. P. 283.

FALLACIES OF FACTUAL SIGNIFICANCE

93

when comprehensive figures are missing, individual cases that happen to
permit quantification are selected and are used without apparent justification
as models of the whole. 74

The same reviewer also criticizes Main's questions for their sim-
plicity, for their failure to incorporate recent and sophisticated work
in social stratification, and for their operational inflexibility. These
deficiencies, I think, generally attach to merely quantitative questions
and can only be corrected by other techniques, which impose methodo-
logical deficiencies of their own which quantification in turn helps to
diminish. The only sensible strategy is a methodological pluralism, which
is necessary to any complex and important project.

A second example of the quantitative fallacy is Richard L. Merritt's
Symbols of American Community (New Haven, 1966). The author's
object is to measure the growth of community awareness in the American
colonies by "content analysis" of seven colonial newspapers in the period
1735-1775. Specifically he counted the number of times words like
"America" and other American symbols appeared, as against the number
of times that "His Majesty's Colonists," "British colonists," or other
British symbols appeared, in articles with British and American datelines.
He reached many conclusions: that the growth of community awareness
was a slow process; that British writers tended to speak of America be-
fore Americans did; that the development of a sense of community pre-
ceded a demand for communitarian integration of institutions; that
individual colony patterns were different in several specific ways but
tended to become more congruent as time passed; that colonial wars
tended to retard the growth of American community awareness, which
was cyclical in nature, like "a typical learning curve" (p. 184).

But all of this is deduced merely from the quantity of British and
American symbols found in the newspapers. He does not sufficiently deal
with important qualitative aspects of his problem, such as the contexts in
which the symbols appeared, or even with the question of whether the
references to Britain or America were favorable or unfavorable in their
connotation. The problem is exceedingly complex — much too complex,
in the present condition of our knowledge, to be resolved in terms of
quantity. A simple "favorable-unfavorable" index would not have sufficed.
Someday, historians may work out ways of using a many-valued symbolic
logic in the quantification of such problems. But at the moment, answers
must be sought in qualitative terms.

By Merritt's strictly quantitative standard of significance, an entirely
false conclusion can obtain. As the author himself suggests, Adolf Hitler

74. Bernard Bailyn, American Historical Review 71 (1966): 1432.

94 INQUIRY

would appear to have had a sense of community with the Jews, according
to a content analysis of Jewish symbols in his speeches and writings.
And the John Birch Society would appear to have a fraternal bond with
the Communist party.

The antinomian fallacy, on the other hand, is the erroneous
idea that facts which count best, count least. Two old-fashioned Amer-
ican historians have publicly committed themselves to this absurdity,
which is more generally assumed than asserted. One of them was Carl
Bridenbaugh, who delivered before the American Historical Associa-
tion a presidential address that was in the nature of a jeremiad. The
professor, like the prophet, pronounced a fearful judgment upon all
the cities of Judah. Among the sins which he singied out for special
condemnation was quantification.

Bridenbaugh allowed some latitude to counting, but not enough.
"The finest historians," he wrote, "will not be those who succumb to the
dehumanizing methods of social science, whatever their uses and values,
which I hasten to acknowledge. Nor will the historian worship at the
shrine of the Bitch-Goddess quantification. History offers radically
different values and methods." 75 Something of the same spirit appears
in a statement by Arthur Schlesinger, Jr.: "As a humanist, I am bound
to reply that almost all important questions are important precisely be-
cause they are not susceptible to quantitative answers." 70

An intense hostility to quantification also appears among radical
historians, who have painfully discovered that counting does not easily
coexist with their ideological preconceptions. An English historian, E.
P. Thompson, bitterly attacks "the orthodoxy of the empirical economic
historians, in which working people are seen as a labor force, as mi-
grants, or as the data for statistical series." His quarrel with quantification
consists in his belief that it tends "to obscure the agency of the working
people, the degree to which they contributed, by conscious efforts, to
the making of history." 77

A more moderate manifestation of the same phenomenon is Bar-
rington Moore's Social Origins of Dictatorship and Democracy (Boston,
1966), which invokes essentialism against statistics. "Statistics," he

75. Carl Bridenbaugh, "The Great Mutation," The American Historical Review 68
(1963): 326.

76. Arthur Schlesinger, Jr., "The Humanist Looks at Empirical Social Research," Amer-
ican Sociological Review 27 (1962): 770.

77. E. P. Thompson, The Making of the English Working Class (New York, 1966),
pp. 12, 196-223.

FALLACIES OF FACTUAL SIGNIFICANCE

95

writes, "are misleading traps for the unwary reader when they abstract
from the essence of the situation the whole structural context in which
social osmosis takes place. As statistics are fashionable now, it is worth-
while stressing this point." 78 Moore is careful to qualify his objections. "I
still have no patience with the machine breaking mentality that rejects
figures out of hand," he writes. "To name this deformation of human-
ist mentality after the Luddites is actually unfair to them; they were
rather more intelligent." 78 But this statement is made in an appendix
called "A Note on Statistics and Conservative Historiography," which
is an obscurantist critique of three important quantitative studies that
reach conclusions contradicting Moore's ideological assumptions. Mr.
Moore comes close to arguing that statistics which sustain his interpreta-
tion of modernization are valid; but statistics which contradict it are
false. 80

But it is among conservative scholars that the antinomian fallacy
is most often apparent, and that an attendant metaphysical bias is most
clearly visible. Historical antinomianism commonly begins with the
assumption that regularities do not exist in history, or that they do not
exist significantly. It holds that every historical event is unique. This
idea, of course, is self -contradictory. If all historical events were unique,
then they would be alike in their uniqueness, and therefore in that respect
they would not be unique. Moreover, they would be alike in their event-
fulness and in their historicity, and therefore not unique in three impor-
tant ways. If every historical happening were sui generis, no language
could be found to communicate its nature. The past would become a
"wilderness of single instances," in Tennyson's phrase. To argue that
every event is unique in its sum and essence is to commit the fallacy of
holism, and the fallacy of essences, which have already been discussed.

"Historical understanding," writes Raymond Aron, "consists of per-
ceiving differences among similar phenomena and similarities among
different ones." 81 The two ideas of similarity and difference must coexist,
if they are to exist at all. Each is meaningless without the other. An
historical interpretation which is cast merely in terms of one, without
the other, is not false but absurd.

78. P. 37.

79. P. 510.

80. Pp. 508-23. The three works which he attacks are D. Brunton and D. H. Pen-
nington, Members of the Long Parliament (London, 1954); G. E. Mingay, "The Size
of Farms in the Eighteenth Century," Economic History Review, 2d ser. 17 (1964):
381-88; and Donald Greer, The Incidence of Terror during the French Revolution
(Cambridge, Mass., 1935).

81. Raymond Aron, "Evidence and Inference in History," in Daniel Lerner, ed., Evidence
and Inference (Glencoe, 111., 1959), p. 27.

96

INQUIRY

Nevertheless, many able historians obsessively persist in the pur-
suit of uniqueness. A prime example is an American historian, Daniel
Boorstin, who bitterly complains that "the rising social sciences have
deprived the historian of his traditional vocation as the high priest of
uniqueness." He continues:

If the historian has any function in the present welter of the social scientific
world, it is to note the rich particularity of experience, to search for the
piquant aroma of life. As contrasted with the abstract, antiseptic dullness
of numbers, "cases," and prototypes [sic]. The historian as humanist
is a votary of the unrepeatability of all experience, as well as of the
universal significance of each human life. Of course the unique cannot
be seen as unique except by reference to a universal (and the word "unique"
itself is such a one), yet there are important practical differences of
emphasis. Today the historian is qualified for a kind of emphasis on
individuality which is too rare in our world. Social scientists are pre-
occupied with what are called the "modal" approach; they are dominated
by the statistical notion of "the category, value, or interval of the
variable having the greatest frequency." They focus upon such concepts
as "status," "personality types," and "occupational mobility" which tran-
scend time and place. And historians who boast of having become inter-
disciplinary are often only confessing that they no longer burn any tapers
before the altar of human uniqueness.' 2

This extraordinary statement is no casual outburst but an extended
commitment, upon which Boorstin is building a major interpretation
of American history in three volumes — a brilliantly suggestive work in
many of its remarkable insights, but at the same time a fallacious attempt
to characterize American culture in terms of qualities which are unique
to it. The result is a history of American religion in which Jonathan
Edwards is an unperson; a history of American politics in which political
parties scarcely exist, and in which Hamilton and Jefferson do not have
a serious difference of opinion; a history of American culture in which
Herman Melville appears mainly as an unsuccessful lecturer; a history
of American regionalism in which the plantation states are exiled to a
distant island of metaphysical alienation, somewhere far below the
southern horizon. 88

Boorstin suggests that a demarcation line might be drawn between
two academic disciplines, with historians responsible for uniqueness and
sociologists, I suppose, responsible for universality. But this surely will
not do. The deficiencies of Boorstin's historical works demonstrate

82. Daniel Boorstin, America and the Image of Europe (New York, 1960), p. 66.

83. Two volumes of Boorstin's trilogy have appeared: The Americans: The Colonial
Experience (New York, 1958); and The Americans: The National Experience (New
York, 1965). The best review which I have seen is Kenneth Lynn's, in Kenyon Review
28 (1966): 116-22.

FALLACIES OF FACTUAL SIGNIFICANCE

97

that the unique and the universal must be combined in each inquiry.
An English historian takes a more tenable position.

No historian really treats all facts as unique; he treats them as particular.
He cannot — no one can — deal in unique fact, because facts and events
require reference to common experience, to conventional frameworks,
to (in short) the general before they acquire meaning. The unique event is
a freak and a frustration; if it is really unique — can never recur in meaning
or implication — it lacks every measurable dimension and cannot be
assessed. But to the historian, facts and events (and people) must be
individual and particular: like other entities of a similar kind, but never
entirely identical with them. That is to say, they are to be treated as
peculiar to themselves and not as indistinguishable statistical units or
elements in an equation; but they are linked and rendered comprehensible
by kinship, by common possessions, by universal qualities present in
differing proportions and arrangements. 84

The fortuitous fallacy is committed by any scholar who ab-
dicates his arduous responsibility of rational selection and allows the
task to be performed for him by time and accident. There is madness in
this method, for it would reduce scholarship to mere sciolism — a smat-
tering of superficial nuggets of knowledge without point or plan or
purpose.

Nevertheless, at least one great popular historian has erected this
method sans method into an avowed standard of selection. He was Giles
Lytton Strachey (1880-1932), English author of Eminent Victorians
(1918), Queen Victoria (1921), Elizabeth and Essex (1928), and
many other volumes which may have reached a larger public than the
work of any other historian in his generation, except Winston Churchill.

Lytton Strachey began his best-remembered work, Eminent Victo-
rians, with the following assertion: "The history of the Victorian Age will
never be written: we know too much about it. For ignorance is the first
requisite of the historian — ignorance, which simplifies and clarifies, which
selects and omits, with a placid perfection unattainable by the highest
art." 85

But the inconvenient presence of excess knowledge did not dismay
Strachey. On the apparent assumption that if ignorance does not exist
it can always be invented, he explored Victorian England as a small
boy plays blindman's buff, covering his eyes and groping over strange
dark objects, and suddenly seizing upon familiar ones.

The careless impressionism of Strachey's method is best communi-

84. G. R. Elton, The Practice of History (New York, 1967), p. 11.

85. Giles Lytton Strachey, Eminent Victorians (New York, 1918), Preface.

98 INQUIRY

cated by his own imagery. "It is not by the direct method of a scrupulous
narration that the explorer of the past can hope to depict that singular
epoch," he wrote. "If he is wise, he will adopt a subtier strategy. He
will fall upon the flank, or the rear, he will shoot a sudden revealing
searchlight into obscure recesses, hitherto undivined." The metaphor
seems more than a little indelicate when we remember that it was
Florence Nightingale's flank which Strachey fell upon, with such un-
seemly, and uncharacteristic, enthusiasm. 8 *

But in all seriousness, such a method is exceedingly ineffectual for
all but the most paltry of popularizing purposes. The reader is diverted
by it, but so is the historian, from his difficult obligation to select factual
statements according to explicit criteria of significance and to tell truths
which are as clear and comprehensive as mortal intelligence will allow.
Any fool can write a readable history, if Strachey's method is all that
is required. Any literary hack can popularize a complicated problem
by this technique, but the nature of that problem cannot be communicated
accurately by it. Such a method guarantees falsehood and gross distortion;
moreover, it prevents the author from knowing how false and distorted
his interpretation actually is.

Lytton Strachey was neither a fool nor a hack. He possessed a rare
and truly remarkable creative gift, and a splendid talent for exposition.
His characters seem so real that the reader thinks they must be true. But
they are merely fantastic inventions.

Strachey's intuitions and expository gifts were perfectly compatible
with a better method of research, which would not have required more
labor than he actually performed. By all accounts, he worked long and
hard in both research and writing. But an improved method would have
taken a toll of Strachey's intentions in one important way. It would have
confined him within limits with which he was apt to be more than a
little impatient — the limits of truth. It would have told him when he was
falsifying, and that was something which Strachey did not wish to hear. 87

Strachey's work had a very great impact, not merely upon the
reading public, but upon the writing of biography as well.

As for Strachey [writes Louis Kronenberger] he was not just a brilliant
biographer, he was a revolutionizing one. An Eminent Victorians had,
indeed, a greater immediate effect than a Ulysses or a Wasteland, for at
one stroke it antiquated a method and enthroned a manner. When Eminent
Victorians appeared, biography while displaying an anarchy of form, was
all too lacking in freedom of speech. Strachey gave it the polish of style,

86. Ibid.

87. For a different interpretation, cf. Michael Holroyd, Lytton Strachey, A Critical
Biography, 2 vols. (New York, 1968), 2:261-305.

FALLACIES OF FACTUAL SIGNIFICANCE 99

the sting of plain speaking, the tensions of fiction, and brought along with
his new kind of technique, a new kind of target. 88

If this statement is more than a little exaggerated, it nevertheless
contains a good deal of truth. After Eminent Victorians there would be
many another biography conceived by the same method, with similar
results. No academic historian has, to the best of my knowledge, openly
and explicitly committed the fortuitous fallacy in print, with the liter-
ary bravado of Lytton Strachey. But there is much of it in the careless,
casual impressionism which still infests much, if not most, historical
writing. Many scholars persist in conducting their research according
to the Strachian spirit, idly falling upon one fascinating fact after another,
without method or control.

This chapter could be extended indefinitely, for there is no fixed
limit to the number of inconsistent standards of significance which his-
torians might adopt. But the nature of history, historiography, and his-
torians is practically such that a few finite forms of error are predominant.
The aesthetic fallacy, the pragmatic fallacy, and the moralistic fallacy
have been deeply rooted in historical writing for 3,000 years. Even the
quantitative fallacy might be traced back to Plato's assertion that God
always geometrizes, or to the Pythagorean idea that all things are num-
bers. Other fallacies are probably transitory, for there are fashions in
error as well as truth. The antinomian fallacy is, I hope, ephemeral. But
other forms of error, equally egregious, will probably take its place.

Nevertheless, there are a few constant and comprehensive rules of
thumb for the determination of sound criteria of factual significance.
First, these criteria must be substantive rather than methodological in
nature. They must be formed from the nature of the subject itself and
from the purpose for which it is studied, rather than from procedures
and techniques. Second, they must be empirical, unlike the pragmatic and
moralistic and aesthetic fallacies. Third, they must be capable of fulfill-
ment, unlike the holist and essential fallacies. Fourth, they must be made
explicitly, for the only alternative to overt criteria of significance are
covert commitments, which are not merely inappropriate but actually
falsifying in their function.

Finally, criteria of significance must not violate what philosophers
call the principle of nonvacuous contrast. There are some words in the
world which are merely meaningless if they are made to stand alone. It
makes no sense to say that all things in the world are short. "Short" is

88. The Atlantic, May, 1968, pp. 85-88.

100 INQUIRY

meaningless unless some things are tall. This reasoning, I think, can be
applied as an operational test to many criteria of significance which his-
torians actually make. It rules out such criteria as uniqueness, and many
others which historians have been industriously applying. Any attempt
to identify all the vacuous contrasts which historians have made would
yield a list which is very long and yet very incomplete. In place of this
strategy, it may be sufficient merely to point out the principle itself.

If false criteria of significance are such as appear in this chapter,
and such as violate these rules, what then might true ones be? The an-
swer, I think, is that a true standard of factual significance is one which is
generated by a sound model of historical explanation.

Historians who seek to frame factual questions and to verify factual
answers almost always do so in order to elicit a historical explanation of
some sort. They rarely do so for the sake of the facts themselves. A his-
torical explanation is an attempt to relate some historical phenomenon
in a functional way to other historical phenomena. Nothing is literally
self-explanatory. An explanation, properly executed, relates the unknown
to the known in a series of orderly inferences. It always explains sentences
— complete sentences. Nobody can hope to have an explanation merely
of "Jeffersonian Democracy," but only of something about Jeffersonian
democracy, something which can be cast in a rounded sentence with a
subject and a predicate.

A fact becomes significant in proportion to its relevance to an ex-
planation model. Historians actually seek to explain things by means of
many different, but sometimes overlapping, explanation models. They
use generalization models, narration models, causation models, motiva-
tion models, composition models, analogy models, and undoubtedly
many others. The nature of these models is very complex. A separate
chapter is therefore assigned to each of them in the following pages.

There is another set of sound standards of factual criteria of sig-
nificance which is not precisely explanatory, but expository and evoca-
tive. A significant fact for most historians is one which helps them to
make a case for their explanation and to communicate its nature to the
reader. The latter partakes of the irrational process which the Germans
call Verstand — a word for which no English equivalent exists. There
cannot be a logic of verstdndliche criteria of significance. But there is a
tacit logic of rational explanation. The next six chapters attempt to specify
it.

PART II

EXPLANATION

CHAPTER IV

FALLACIES OF
GENERALIZATION

What distinguishes the historian from the collector of
historical facts is generalization.

— E. H. Can-
Pour faire de l'histoire, il faut savoir compter. [To do his-
tory, one must know how to count.]
— Georges Lefebvre

Scholars still solemnly engage in controversies over questions such as
"Should a historian generalize?" One might as well ask "Should a his-
torian use words?" Generalizations are embedded in his language, in his
thought, and in his explanation models. There are, of course, other
modes of explanation. But none is more common, or more commonly
abused, than generalization.

The term "generalization," in ordinary usage, means everything
and nothing. In a recent discussion of the problem, an attentive critic
found nine distinguishable conceptions of generalization:

1. "Labeling," or "classificatory," concepts; e.g., "feudalism."

2. Universal "laws" or regularity statements; e.g., "Ideas follow
trade routes."

3. "Limited," or summative, general statements, specifying the
conditions that obtained in a particular geographical area or during a
determinate period of time; e.g., "Successful revolution occurred [in the
eighteenth century] only where the agricultural population generally
collaborated with middle-class leaders."

4. Statements asserting the existence of a trend or tendency; e.g.,
"The Speaker was a power in the House, but as the Elizabethan period
went on, his power was on the wane."

5. Statistical regularities.

104

EXPLANATION

6. General characterizations of some particular historical figure
with regard to his motivation, activities, etc.

7. Particular explanations or interpretations of events.

8. Evaluative assessments.

9. Procedural rules for the selection of historical material and data,
the authentication of evidence, etc. 1

In this chapter, generalization will refer to statements of statistical
regularity, a category which cuts across all others on the list. A statistical
generalization is a descriptive statement which is inferred from particular
facts by a special process of reasoning. It is often built into a complex
explanation model with other components and is sometimes employed as
an explanation in its own right. To explain is, in common usage, merely
to make plain, clear, and understandable. This is precisely what a
statistical generalization does. It explains what, how, when, where, and
who. It does not explain why, but often we don't need or wish to know
why.

Historical science presently hovers naked and trembling on the
edge of quantification, with Clio in the huddled, hesitant posture of
September Morn. But it is already irretrievably committed to the logic
of statistical reasoning, in many quasi-quantitative ways. Statistical
generalizations are implicit not merely in impressionistic adjectives like
"some" and "most" and "many," but also in nouns like "majority" and
"democracy" and verbs like "expand" and "grow." A single term can
serve to summarize an implicit statistical statement; a single impression-
istic sentence can contain an appalling number of implicit statistical
series.

It is still uncommon for a historian to make explicit and formal use
of statistical reasoning. Consequently, the statistical fallacies which his-
torians commonly commit are, at present, of an elementary and transi-
tional nature. They often consist in a confusion of quantitative and im-
pressionistic procedures. As practices change, historians will probably
learn to make more sophisticated blunders. But in the meantime, our
business is with crude low-order mistakes of the sort which commonly
occur in everyday thought. All of them are procedural errors, by which
false inferences are derived from true particulars.

Fallacies of statistical sampling occur in generalizations
which rest upon an insufficient body of data — upon a "sample" which

L Patrick Gardiner, reviewing Louis Gottschalk, ed., Generalization in the Writing of
History and Theory 3 (1965): 351.

FALLACIES OF GENERALIZATION 105

misrepresents the composition of the object in question. Sound methods
of sampling have been satisfactorily discussed at length by statisticians
with a degree of clarity and accuracy which no historian can hope to
attain. 2 But when it comes to unsound sampling, historians qualify as
experts, by reason of their long experience in the ways of error. Rarely
have so many generalized so much from so little as have impressionistic
historians in the past generation.

One example is an assertion, sometimes made by historians of
the framing of the United States Constitution, that there was a signifi-
cant age difference between Federalists and Antifederalists. A recent
statement of this thesis is by Stanley Elkins and Eric McKitrick, in
The Founding Fathers: Young Men of the Revolution (Washington,
1962). Elkins and McKitrick attempted to develop a generational
interpretation of the fight over the Constitution as an alternative to
Charles Beard's battered economic thesis. These two historiographical
knights errant boldly attacked the problem, as Beard himself said of
another scholar, sans fear and sans research. Their argument rests upon
a so-called sample of nine Federalists and nine Antifederalists:

Federalists Antifederalists

Robert Morris, b. 1734 Samuel Adams, b. 1722

John Jay, b. 1745 Patrick Henry, b. 1736

James Wilson, b. 1742 R. H. Lee, b. 1732

Alexander Hamilton, b. 1755? George Clinton, b. 1739

Henry Knox, b. 1750 James Warren, b. 1726

James Duane, b. 1733 Samuel Bryan, b. 1750?

George Washington, b. 1732 George Bryan, b. 1731

James Madison, b. 1751 George Mason, b. 1725

Gouverneur Morris, b. 1752 Elbridge Gerry, b. 1744

The authors conclude that "the age difference between these two groups
is especially striking. The Federalists were on the average ten to twelve
years younger than the Anti-Federalists." The arithmetic is slovenly —
the average (mean) difference is in fact 9.8 years. But more serious
is the validity of the sample, which is absurdly small and extremely
biased.

It is easy to invent a different list of equal length which would sup-
port an opposite conclusion.

2. For a good introduction see W. Allen Wallis and Harry V. Roberts, Statistics, A
New Approach (New York, 1956), chap. 11, passim.

106

EXPLANATION

Federalists

Antijederalists

Robert Morris, b. 1734
James Duane, b. 1733

Patrick Henry, b. 1736
George Clinton, b. 1739
Elbridge Gerry, b. 1744

George Washington, b. 1732
Benjamin Franklin, b. 1706
James Madison, b. 1 75 1
Nathaniel Gorham, b. 1738

John Francis Mercer, b. 1759
John Lansing, Jr., b. 1754
Benjamin Austin, b. 1752

John Rutledge, b. 1739
Roger Sherman, b. 1721
W. S. Johnson, b. 1727

Luther Martin, b. 1748?
Samuel Bryan, b. 1750?
Melancton Smith, b. 1744

These Antifederalists are, on the average, 13.9 years younger than the
Federalists. Adding the Elkins-McKitrick arithmetical inflation quo-
tient, we conclude, "The age difference between these groups is es-
pecially striking. The Federalists were, on the average, 14 to 16 years
older than the Antifederalists."

A more comprehensive study, by Jackson T. Main, suggests that
both lists are invalid. Working from a much larger sample, he found
no significant age difference between friends and foes of the Constitu-
tion. "In Pennsylvania the Antifederalists were older by . two years,"
he writes, "but in New York there was no difference at all; in South
Carolina the median age of eleven Antifederalists was 33; of thirty-two
Federalists, only 29; in Massachusetts, the average Antifederalist
was 52, the Federalist 51. All told, the Federalists were about two
years younger than their antagonists. It is hard to see how this could
have made any difference." 3 Other generational interpretations of the
same problem might be entertained. It is conceivable that extremists in
both camps were younger than moderates; and Westerners were younger
than Easterners. But the generational lines appear to have crossed party
lines at right angles, // Main's work is correct. (Maybe it isn't.)

The probability of a sampling error tends to diminish as the size
of the sample increases. But size alone is no protection. The classic
example was a massive effort by the Literary Digest to forecast the
presidential election of 1936. More than 10,000,000 ballots were sent
out. Something like 2,367,523 came back, mostly marked for Alf
Landon. The poll predicted 370 electoral votes for the Republican
candidate, and 161 for Roosevelt. In the real election, Roosevelt won
523 votes, Landon 8. What went wrong? The Digest, it seems, sent

3. Jackson T. Main, The Antifederalists: Critics of the Constitution (Chapel Hill,
1961), p. 259. Main's conclusions may in turn be corrected by other inquiries, now in
progress.

FALLACIES OF GENERALIZATION

107

ballots to addresses collected from the subscription lists of magazines
and also from telephone directories and automobile registration lists.
But magazines, telephones, and automobiles were not randomly dis-
tributed among the American population in 1936, as many will ruefully
recall.

This famous statistical disaster was not the only sampling error
which was made during the presidential campaign of 1936. In October
of that year, a Republican leader in Chicago discovered that demands
for Alf Landon buttons greatly exceeded those for the Democratic em-
blems. On this basis, he wrote to Landon's headquarters, confidently
predicting that the Kansas governor would carry the city. But a demand
for Landon buttons was not equivalent to a demand for Landon. The
Republicans had distributed a gorgeous badge for their candidate — a
brown tin disk set against a yellow felt background, in the image of a
Kansas sunflower. For the possession of this splendid trinket, many a
red-blooded American boy in 1936 might cheerfully have sold his
mother into slavery. According to one learned scholar, the exchange
rate among small boys in Chicago was two or three plain, ordinary
Roosevelt buttons to one of Landon's. 4 Most Americans sensibly chose
the best from both parties — the Republican campaign badge and the
Democratic candidate.

There are many other examples of sampling errors in historical
scholarship, some of which are explicitly statistical and others implicitly
so. One scholar, Edward Pessen, examined election returns in Boston
during the Jacksonian era. He concluded that workingmen more gen-
erally voted Whig than Democratic and that they either rejected or
ignored so-called workingmen's candidates. But another scholar sharply
criticized Pessen for failing to "use some tool of analysis that will
neutralize the effect of the general level of voting in the area and allow
us to see to what degree voting Whig or Democratic is related to the
socio-economic nature of the wards." Working from the same sources
by a different and more defensible method, the critic concluded that
"Jackson and his political allies did get their support from working
class groups," and that there was a good inverse correlation between
votes for Democratic and workingmen's candidates and an economic
index. 5

Another case of an explicitly statistical sampling error appears in
Robert E. Brown's attempt to show that something which he calls a

4. Donald R. McCoy, Landon of Kansas (Lincoln, 1966), p. ix.

5. Edward Pessen, "Did Labor Support Jackson?: The Boston Story," Political Science
Quarterly 64 (1949): 262-74; Robert T. Bower, "A Note on 'Did Labor Support
Jackson?: The Boston Story,"' ibid., 65 (1950): 441-42.

108

EXPLANATION

"middle class democracy" existed in pre-Revolutionary Massachusetts
and that property was widely and equitably distributed, and the fran-
chise as well. But Brown's thesis is much exaggerated in both respects,
and the exaggeration is partly sustained by serious errors in the sampling
of probate records, tax lists, and voter lists. 6

Other sampling errors are implicitly, rather than explicitly, sta-
tistical. They occur in simple impressionistic statements, such as the
following observation by the late V. O. Key on the alleged suicidal
tendencies of Southern literati: "A depressingly high rate of self-destruc-
tion prevails among those who ponder about the South and put down
their reflections in books," he wrote. "A fatal frustration seems to come
from the struggle to find a way through the unfathomable maze formed
by tradition, caste, race and poverty." 7

A friend of Key's asked how many suicidal Southern writers he
could name. There were only three — W. J. Cash, Clarence Cason, and
another whose name Key couldn't recall. 8 Some Southern historians may
consider W. J. Cash to be a host within himself, but not for the pur-
poses of quantification.

Impressionistic sampling errors are sometimes so gross than no-
body is taken in by them — probably not even the author. A Russian
historian, B. I. Bukharov, has used opinions expressed in the Daily
Worker as representative of American opinion generally. 9 But other
errors can be very subtle and insidious. The difficulties entailed in
statistical sampling have sometimes tended to make statistics generally
suspect in the eyes of some historians, more than a few of whom are
prepared to accept Disraeli's famous statement that there are lies,
damned lies, and statistics. But statistics themselves do not lie — only
statisticians, and if they lie, they are apt to lie to themselves, in
Dostoevski's phrase. And their confusions are clarity itself compared to
the impressionistic muddles which so many historians get themselves
into by an even more dangerous method.

With skill, practice, and a little luck (impressionists need a lot
of luck), statistical sampling can be very reliable indeed. Anglo-
American intelligence officers in World War II used a sampling method
to estimate German production figures from serial numbers on cap-

6. Robert E. Brown, Middle-Class Democracy and the Revolution in Massachusetts,
1691-1780 (Ithaca, 1955); John Cary, "Statistical Method and the Brown Thesis on
Colonial Democracy," William and Mary Quarterly, 3d ser. 20 (1963): 251-64, with
a rebuttal by Brown, 265-76.

7. V. O. Key, Southern Politics in State and Nation (New York, 1947), p. 664.

8. Joseph L. Morrison, W. J. Cash: Southern Prophet (New York, 1967), p. 140.

9. See Ernest May's review in American Historical Review 67 (1961): 87.

FALLACIES OF GENERALIZATION

109

tured equipment. Their estimates are alleged to have been more ac-
curate than those of the German government itself, and more quickly
available than inventory estimates. Many historians, still wedded to
the whole truth, will find themselves at a disadvantage comparable to
that of the Germans. To repudiate samples and to count everything is
an arduous undertaking which is not only unnecessary in many cases
but undesirable as well. The incompleteness of historical evidence is
itself an argument for scientific sampling, rather than an argument
against it.

$»» The fallacy of the lonely fact is the logical extension of a
small sample, which deserves to receive special condemnation. It may be
defined as a statistical generalization from a single case. There is a
story, perhaps apocryphal, of a scientist who published an astonishing
and improbable generalization about the behavior of rats. An incredulous
colleague came to his laboratory and politely asked to see the records
of the experiments on which the generalization was based. "Here they
are," said the scientist, dragging a notebook from a pile of papers on
his desk. And pointing to a cage in the corner, he added, "there's the
rat."

There are many astonishing and improbable generalizations in
historical scholarship in which the critical reader will smell a singular
rat. As long as the majority of historians continue to conduct their
"research" impressionistically and to cast their findings in a simple
narrative, the fallacy of the lonely fact is likely to flourish. A special
stylistic device has developed around it. Whenever the reader sees a
mighty generalization, followed by a minute example, and the telltale
phrase "for instance," or "for example," he should be on his guard
against this error.

But often the fallacy of the lonely fact occurs without warning.
The only defense is research in depth, of the sort which readers are
rarely equipped to carry out. In the past half century, many (perhaps
most) interpretations of the thought and conduct of the members of the
Federalist party have derived primarily from the case of a single ec-
centric West Indian — Alexander Hamilton. In the past twenty years,
a bold revisionary trend has set in. Textbook writers and political
scientists now generalize upon the Federalists from two cases, an ec-
centric West Indian and an eccentric New Englander — Hamilton and
John Adams. Similarly, many broad generalizations about the Jeffer-
sonian party are precariously based upon the behavior of Thomas

110

EXPLANATION

Jefferson himself. But as we learn more about the politics of the early
republic, Jefferson is beginning to appear as a very special sort of Jeffer-
sonian. These three statesmen — Hamilton, Adams, and Jefferson — were
exceptional men in more senses than one. Each of them, in his own way,
was sui generis. This is not to argue that they had nothing in common
with their political colleagues. But common qualities can never be located
by investigation of themselves alone.

The most successful recent survey of American political history,
Richard Hofstadter's The American Political Tradition and the Men
Who Made It (New York, 1948), contains many examples of this
fallacy. The book is a sequence of single facts, punctuated by sweeping
generalizations. Jefferson, Jackson, Calhoun, Lincoln, Wendell Phillips,
Bryan, Wilson, Hoover, and the two Roosevelts are each individually
sketched as representative of "main currents of American political
sentiment." 10 In only two chapters out of twelve is a different strategy
developed. Hofstadter's book surely deserves its splendid reputation —
what it lacks in the breadth of its empirical base, it more than matches
in the depth of its remarkable intuitions. But had the author's intuitive
excellence been wedded to the discipline of a better method, a very
great book indeed would undoubtedly have been the result.

Other specimens of this fallacy have occurred in the field of pre-
history, where archaeologists and historians have generalized broadly
from a single find. Some of them concluded from a skeleton found at La
Chapelle-aux-Saints that Neanderthal man had the posture of a pretzel
and the locomotion of a crab — that "the head was so balanced on the
spine that it hung forward; the structure of legs and feet permitted only
a shuffling gait." But re-examination of the same evidence has suggested
that the individual in question had a bad case of arthritis, and that
normal Neanderthal men stood straight and strode confidently in pur-
suit of their prey. 11 It is as if an archaeologist, a millennium hence,
were to generalize upon the stature of twentieth-century Homo ameri-
canus from the Brobdingnagian bones of Wilt-the-Stilt Chamberlain.

The fallacy of statistical special pleading occurs whenever an
investigator applies a double standard of inference or interpretation to
his evidence — one standard to evidence which sustains his generaliza-
tion and another to evidence which contradicts it.

Imagine a scholar who hypothesizes that Republican ladies tend
to have blond hair, while Democratic females are brunette. He carefully

10. P. x.

11. Gordon Childe, What Happened in History, new ed. (Baltimore, 1964), p. 40.

FALLACIES OF GENERALIZATION

111

conducts his research at two national conventions, and when he tabu-
lates his results, he discovers to his horror that they are:

Blondes Brunettes Other

Republican ladies 20 40 10

Democratic ladies 20 40 10

But our scholar is not so easily discouraged. He proceeds to the second
stage of his inquiry, which is called "interpretation" of the evidence.
First, he re-examines his twenty blond Democrats, discovers that ten
of them dyed their hair, and shifts them to the brunette column. Second,
he studies the forty brunette Republicans, discovers that twenty-five
of them had been blond in childhood, and shifts them to the blond
column. Third, he takes a look at the miscellaneous column, which in
each party includes five ladies with red, white, gray, etc., hair, and five
ladies with dark-blond or light-brown hair. The latter, if Republican,
he classifies as dark blond; but if Democrat, light brunette. He adds the
dark blondes to the blond column, and the light brunettes to the brunette
column. The results read:

Blondes Brunettes Other

Republicans 50 15 5

Democrats 10 55 5

He works out his coefficients of correlation and triumphantly pub-
lishes the results in The American Behavioral Scientist. None of
these manipulations necessarily involves an outright fraud. It is possible
that our scholar did not deliberately fudge any of his figures. He may
have been utterly unaware of his own errors.

There are many examples of this fallacy in actual historical prac-
tice. One of them is David Donald's The Politics of Reconstruction
(Baton Rouge, 1965). Donald tried to explain why Republican con-
gressmen in the Thirty-ninth Congress (1866-1867) voted as "radicals"
or "moderates" on legislation before them. He hypothesized that the
difference between them lay not in "personality, ideology, geographical
origins, or social and economic status" but rather in their political
situation — specifically, that Republicans who held safe seats — with
wide electoral margins — tended to be radical, while Republicans who
were elected by narrow and precarious pluralities tended to be moderate.
Unfortunately for Donald's hypothesis, the result of his research showed
that margins of electoral victory for both radicals and moderates were

112

EXPLANATION

almost identical — 59.3 percent to 59.2. But Donald proceeded to
"interpret" his evidence until it conformed to his hypothesis. He argued
that in some states, where there was an allegedly powerful Republican
organization, a majority of 52 percent was security itself for a radical,
but that in another state, a majority of 58 percent was too small to serve
the same purpose. One qualification was neaped upon another with an
industry and an ingenuity which were worthy of a better cause.

To maintain his propositions [one reviewer wrote] Donald is forced to
resort to a "cake and eat" approach. A Radical's election by a large vote
is prima facie evidence of a safe seat. But when Moderates take office with
large margins Donald introduces the realm of psychology, that is, his
guess as to how secure the congressman felt. . . . There is no shortage of
ingenious explanations here, but there seems to be little sense in accumulat-
ing data only to shuffle and disregard it. 12

A second and more impressionistic example is a recent work by
Alan Heimert. In Religion and the American Mind from the Great
Awakening to the Revolution (Cambridge, 1966), he argued the thesis
that there was a correlation between the religious divisions between
Old Light opponents of the Awakening and New Light friends of it,
on the one hand, and the political cleavage between friends and enemies
of the Revolution. This interpretation is supported not with quantita-
tive evidence but with qualitative interpretations of thought and ex-
pression. But Heimert applies a double standard of inference and
interpretation to his evidence. He is often forced to argue that his his-
torical subjects really meant the opposite of what they said.

One critic has complained,

For Professor Heimert things are seldom what they seem, but if we know
which side of the cleavage a man stands on, we need have no difficulty in
penetrating the disguises of reality. . . . The world he offers us has been
constructed by reading beyond the lines of what men said; and what he
finds beyond the lines is so far beyond, so wrenched from the context, and
so at odds with empirical evidence, that his world, to this reviewer at
least, partakes more of fantasy than of history. 13

Maybe this judgment is a little on the harsh side, but I think it is
roughly correct. Heimert has so seriously overstrained his interpreta-
tion by an unfortunate — and certainly unintentional — double distortion
of his evidence that it is quite impossible to know how much credence
to put into his thesis until another scholar undertakes a more careful
analysis of the same problem by means of a better method. Nothing

12. David Rothman in Political Science Quarterly 81 (1968): 334-36.

13. Edmund S. Morgan, in William and Mary Quarterly, 3d ser. 24 (1967): 454-59.

FALLACIES OF GENERALIZATION

113

is more pathetic than so many years of labor utterly wasted and undone
by a procedural fallacy of this sort.

It would, of course, be a great mistake to assume that statistical
special pleading is a sign of fraud whenever it appears. It is extraor-
dinarily easy for a historian to distort his evidence in this way without
ever intending consciously to do so. A great English economic historian
of an older generation, Sir John Clapham, once observed, "It is very easy
to do this unawares. Thirty years ago I read and marked Arthur
Young's Travels in France, and taught from the marked passages. Five
years ago I went through it again, to find that whenever Young spoke
of a wretched Frenchman I had marked him, but that many of his
references to happy or prosperous Frenchmen remained unmarked." 14
The same thing has happened in almost everybody's work.

The fallacy of statistical impressionism occurs whenever a his-
torian casts an imprecise, impressionistic interpretation into exact num-
bers. The result tends to be speciously specific or specifically specious.

An example of a historian who casts impressions into numbers is
Lee Benson, in The Concept of Jacksonian Democracy: New York as
a Test Case, 2d ed. (New York, 1964). Benson's book is both a polemi-
cal plea for quantification in history, and an intensive analysis of the
New York elections of 1844 as a case study in political ecology. It is,
I think, more successful as a manifesto than as a monograph, for its
major points are either inaccurate or unsubstantiated. The reader finds
much of the rhetoric of quantification in Benson's book but little of its
reality. The Concept of Jacksonian Democracy combines the worst of
both methodological worlds — history is researched as an art and written
as a science.

Chief among the theses of this controversial book is an argument
that there can be no clear and simple correlation between party affilia-
tion and economic wealth, in the New York elections of 1844 in par-
ticular, and American political history in general. Instead, Benson sug-
gests a six-part multivariate analysis of voting patterns in terms of pre-
vious voting behavior, economic groups, ethnocultural groups, religious
groups, residential groups, and regional groups.

Benson's thesis is asserted rather than proved. One looks in vain,
through 340 pages crowded with numbers, footnotes, and tables, for
any comprehensive and unequivocal attempt to calculate the correlation

14. Quoted in Edward P. Thompson, The Making of the English Working Class (New
York, 1964), p. 210.

114

EXPLANATION

of political affiliation with economic wealth. Instead, there are declara-
tive sentences punctuated with impressionistic scraps of statistics such
as the following:

Perhaps the most dramatic way to illustrate the inaccuracy of traditional
claims about voting behavior is to present the data for the two towns in
the state that received the highest and lowest Democratic vote. Clarkstown
in Rockland County was a "very prosperous" ($629) Hudson River town,
but its 89.9% Democratic vote established it, by a considerable margin,
as that party's banner unit. The second highest Democratic unit (83.9
per cent) was an extremely "poor" ($494), isolated interior town (Croghan
in Lewis County). The lowest Democratic unit was Putnam, a "marginal"
town ($339) in Washington County, where the Democratic vote was
18.5 per cent of the total. 16

These impressionistic figures lend a sense of statistical authority
to Benson's book, but little substance. Another inveterate historical
quantifier has wisely observed that "to support an argument by a few
examples, though it may be a persuasive rhetorical device, is not logi-
cally adequate. There are exceptions to most historical generalizations,
and, if the citation of occasional instances were accepted as proof, it
would be possible to prove almost anything." 18

Though Benson asserts, or implies, that he systematically com-
pared economic and political data throughout the state, he strangely
omits anything in the way of a comprehensive statistical generalization
which bears upon that critical question.

He comes close twice. There is only one table, out of many in
the book, which displays the two variables side by side for any area of
the state — a tabulation of towns in Delaware County. This table shows
the average valuation of dwellings in eighteen towns, the percentage
of the Democratic vote in sixteen of them, and their rank order in both.
But there is nothing at the bottom of the table, in the way of a summary

15. P. 150. The dollar values refer to the average (assessed) value of dwellings per
family.

16. William O. Aydelotte, "Quantification in History," The American Historical Review
71 (1966): 805. Another quantifier has commented directly on Benson's book, "Despite
his concern for theoretical explication, Professor Benson's work sometimes falls short
of the standards that many behavioural scientists consider essential. One searches the first
edition of The Concept of Jacksonian Democracy in vain for any detailed discussion of
the methods by which he selected his indicator precincts, or of the numbers of voters in
his sample, or of correlations or significance tests underlying the party preference per-
centages which he ascribed to the various ethno-cultural groups living in New York
during the 1830s and 1840s." Allan G. Bogue, "United States: The 'New' Political His-
tory," Journal of Contemporary History 3 (1968): 11. In the second edition, Benson
added a descriptive methodological statement which appears to have satisfied Bogue.
But an explicit description of a bad method does not thereby convert it into a good one.
The fatal flaws remain.

FALLACIES OF GENERALIZATION

115

or a statistical generalization, and nothing in the text but statistical
snippets.

I took the sixteen towns for which both economic and political
data appear in the table and tried three experiments: First, a simple
two-by-two table, in which the towns were equally divided on two axes:
the eight most Democratic, the eight least Democratic, the eight most
wealthy, and the eight least wealthy. The results were:

8 most 8 least

wealthy wealthy

8 most
Democratic

8 least 6
Democratic

This appears to contradict Benson's thesis and also to undercut the im-
pressionistic examples which he extracted and discussed in his text.

Next I averaged the wealth of the towns which the Democrats
carried by more than 50 percent ($331) as against those in which
they received less than that percentage ($414). The results once again
contradicted Benson's assertion that "no significant relationship existed
between wealth and voting in 1844." Finally, I calculated a Spearman
rank correlation coefficient for the sixteen towns and got a positive
correlation of .3, which leaves the issue in doubt. That marginal figure
neither confirms nor refutes Benson's sweeping generalization but requires
further evidence of the sort which he does not provide. 1 '

A second piece of evidence which appears to bear directly on the
issue of wealth and politics is Benson's analysis of a directory called
Wealth and Biography of the Wealthy Citizens of New York City,
Comprising an Alphabetical Arrangement of Persons Estimated to Be
Worth $100,000, and Upwards — With the Sums Appended to Each

17. The disparity of results obtained by these three different statistical tests may serve
to suggest the need for historians to study the mathematical assumptions embodied in
the statistical tools at their command. Lancelot Hogben has remarked upon the "feverish
concern of biologists, sociologists and civil servants to exploit the newest and most
sophisticated statistical devices with little concern for their mathematical credentials
or for the formal assumptions inherent therein." Historians are beginning to use the
oldest and least sophisticated statistical techniques in precisely the same uncritical
spirit. Lancelot Hogben, Statistical Theory: The Relationship of Probability, Credibility
and Error (New York, n.d.) p. 13.

116

EXPLANATION

Name, published by Moses Beach in 1845. 18 Benson examined a sample
of a few names and concluded that Democrats and Whigs appeared in
almost equal numbers. But another scholar has recently studied the
whole list in a comprehensive fashion and discovered a very close and
symmetrical correlation between wealth and Whiggery. 19

Benson's methodological misdemeanors are not a mark of depravity.
They are not a proof of the inherent inadequacy of quantification in
history. The remedy is not less quantification but more. And Benson
himself has, I think, done more than any other single individual to make
possible more thorough and exact quantification in American political
history. His work in the organization of the American Historical Asso-
ciation's program for the computer programing of political and social
data, a program which is already sufficiently advanced to be useful and
even indispensable to American political historians, is a giant step
forward. But The Concept of Jacksonian Democracy was in some re-
spects a step back.

The fallacy of statistical nonsense serves as a catchall category
for a miscellany of statistical mumbo jumbo, all of which has one quality
in common: it is, in context, literally meaningless. Sometimes it consists
in a statistical generalization which may be accurate and important in
some other context, but which is irrelevant to the problem at hand. A
historian has argued, for example, that a measure of the effectiveness
of Communism in the U.S.S.R. appears in the fact that "the number of
doctors in Russia had increased from 1,380 in 1897 to 12,000 in
1935. " 20 But the meaningful comparison for this problem is the increase
in the number of doctors from 1917 to 1935. The Soviet regime has
often received statistical credit for economic and social development
at the end of the czarist period.

Another example of the same sort of error is from American
political history, where one scholar attempted to measure the degree
of party cohesion in the first session of the Second Congress by recording

18. Reprinted in Henry Wysham Lanier, A Century of Banking in New York: 1822-
1922 (New York, 1922), pp. 151-84.

19. Frank O. Gatell, "Money and Party in Jacksonian America: A Quantitative Look
at New York's Men of Quality," Political Science Quarterly 82 (1967): 235-50. There
are several signs of trouble in Gatell's essay, too. I compared the Beach list with other
overlapping directories for Brooklyn and found that old established wealth tended to
be more completely represented than new wealth, and wealthy citizens of English and
Dutch descent more than Irish wealth, such as it was. This suggests that the Beach list
is skewed toward Whiggish wealth, and that Gatell's conclusions require qualification.
But they demonstrate a gross and serious inaccuracy in Benson's use of the Beach list.

20. Bernard Pares, Russia (New York, 1949), p. 137.

FALLACIES OF GENERALIZATION

117

the number of times the votes of each congressman agreed or disagreed
with the vote of James Madison. The figures are accurate, but they
are not an appropriate test of party regularity, for Madison himself was
something of a swing-voter at this time. Moreover, to categorize party
patterns in these terms is to permit only two patterns to emerge — a pat-
tern of consistent divergence from Madison's record and one of agree-
ment with it. It would exclude the discovery of multiparty patterns. 21

The second general category of nonsensical statistical generaliza-
tions includes statements which float in an interpretative vacuum, with-
out an adequate control group as a reference point. Thus, David Donald
attempted a statistical analysis of abolitionist leadership and found that
the median age was twenty-nine, that 85 percent came from the North-
east, 60 percent from New England, and 30 percent from Massa-
chusetts; that there was a preponderance of Congregationalists and
Presbyterians and Quakers; that most were from respectable families,
neither rich nor poor; that few were immigrants; that few were born
in the city; and that few had any connection with industrial activity.
These statistical generalizations are not, however, weighed against
national norms, or the leadership of other movements, or any other
control. Without such a control, they are largely empty of meaning. 22

The third general category of statistical nonsense consists in shift-
ing statistical bases as a correlation is run. James Sterling Young, in
The Washington Community, tried to prove that there was a pattern
of residential segregation in the District of Columbia during the Jeffer-
sonian era, with congressional legislators living in one community and
executive officers in another. But he compared legislative residences in
one period (1807, apparently) with executive residences in another
(1800-1828, apparently mostly toward the end of this period), and
both in turn with a control from a third period (1800). All of this was
at a time when Washington was being built, and residential patterns were
changing, and the changes cut against the author's thesis. 23

Other examples of comparable errors appear in statistics text-
books, which also supply illustrations of a fourth category of statistical
nonsense, not much in evidence in historical writing — yet. But it is
one which is likely to become more common as quantification becomes
more routine. This fallacy of the fourth type consists in a shifted statis-

21. Noble Cunningham, The Jeffersonian Republicans: The Formation of Party Organi-
zation, 1789-1801 (Chapel Hill, N.C., 1957), pp. 22, 267-72.

22. David Donald, "Toward a Reconsideration of the Abolitionists," Lincoln Recon-
sidered (New York, 1956), pp. 19-36.

23. James Sterling Young, The Washington Community 1800-1828 (New York, 1966),
pp. 67, 69.

118

EXPLANATION

tical method, which has the same results as a shifted statistical base.
It includes a variety of technical errors of various degrees of sophistica-
tion, in which medians, modes, and means are mixed up, or averages
of any sort are confused with ranges of dispersion, and stupid, careless
mistakes in which "natural" and "common" logarithms are mistaken
for one another, or one unit of measurement is substituted for another. 24
Historiographical examples have, however, already begun to appear.
In Terry G. Jordan's German Seed in Texas Soil: Immigrant Farmers in
Nineteenth Century Texas (Austin, 1966) there is an attempt to com-
pare statistically the condition of German and Anglo-American farmers.
But the author relies principally on averages, without applying informa-
tion as to ranges and dispersions and medians, all of which are func-
tional to his thesis.

$»» Fallacies of statistical probability are numerous, and some-
times very technical, and as yet rarely committed in historical scholar-
ship. We might, however, briefly note two of them. The first is the
error of assuming that the most probable distribution will occur, exactly,
in any given instance. Abraham Kaplan writes,

This point played an historic part in the Dreyfus trial, where the prosecution
argued that Dreyfus' correspondence must be in code because the frequency
distribution of the letters of the alphabet contained in the correspondence
deviated from what is "normal" for the French language. The testimony
of Poincare for the defense that the most probable distribution is highly
improbable was not very convincing, in spite of its being correct. (Possibly
a contributing factor was that Poincare had identified himself on the stand
as the greatest living expert on probability, a tactical error which he later
justified to his friends by pointing out that he was under oath at the
time.) 25

A second and related simple form of probability error is the familiar
gambler's fallacy, which is the error of assuming that if a coin has
come down heads one time, it is more likely to come down tails the
next, because the probability (assuming the coin to be symmetrical) is
.5 for heads and .5 for tails. But the "law of averages" does not work
this way. The odds are always even in every single toss.

24. Darrell Huff, How to Lie with Statistics (New York, 1954); Ernst Wagemann,
Narrenspiegel der Statistik (Bern, 1950); J. B. Cohen, "The Misuse of Statistics," Journal
of the American Statistical Association 33 (1938): 657-74; W. Allen Wallis and Harry
V. Roberts, Statistics: A New Approach (New York, 1956), chap. 3; Hans Zeisel, Say
It with Figures, rev. 4th ed. (New York, 1957), passim; M. J. Moroney, Facts From
Figures, rev. 3d ed. (Baltimore, 1956), passim.

25. Abraham Kaplan, The Conduct of Inquiry (San Francisco, 1964), p. 224.

FALLACIES OF GENERALIZATION

119

$»» The ecological fallacy is a false form of statistical inference
which is common in quantitative research. Its popularity is sustained
by the fact that it appears to be a shortcut to a significant statistical
conclusion. It is a form of error which is more easily committed than
characterized, and most clearly explained by an example.

W. S. Robinson, who discovered and named this fallacy, provides
a good illustration. He attempted to estimate the statistical relationship
between illiteracy and race (white-Negro), by three statistical methods.
First, he correlated marginal percentages between Negroes and illiteracy
in nine standard census regions of the United States. The result was a
positive Pearson correlation coefficient of .946. Next, he correlated per-
centages of Negroes and illiterates in the forty-eight states. This time
he obtained a different estimate — a positive Pearson coefficient of
.773. Finally, he attempted to run a correlation, not between ecological
areas, but between individuals — white and Negro, literate and illiterate.
The result was still a third figure: a low positive Pearson coefficient of
.203. The ecological fallacy consists in a confusion of the first two
ecological correlations with the third individual correlation. 28

An individual correlation is one between discrete and indivisible
objects — namely, people, in this case. An ecological correlation is
between percentage figures which refer to classes of individuals — areal,
or geographical or ecological classes. The ecological fallacy occurs
when classes are used which do not coincide with the variable being
measured. The degree of error is regressive, as the classes increase in
number and diminish in size, and also in proportion with the degree of
coincidence with the variable under investigation. Suppose, for instance,
that General Sherman had his way, and Negroes were entirely segregated
by region from whites in the United States. Suppose all Negroes lived
in census region one, and all whites in regions two through ten. Then,
an ecological correlation should yield the same results as an individual
correlation. 26 * But not otherwise. It is possible, by the use of regression
techniques, to derive individual correlations from ecological data. 27
But it is incorrect to use an ecological correlation itself as an inter-
changeable substitute for an individual correlation. This point, im-
portant as it is for sociologists, is still more important for historians
who seek to use quantitative materials. If Robinson, in his example,

26. W. S. Robinson, "Ecological Correlations and the Behavior of Individuals," Ameri-
can Sociological Review 15 (1950): 351-7; Leo Goodman, "Ecological Regression and
the Behavior of Individuals," and "Some Alternatives to Ecological Correlation," in
ibid. 18 (1953): 663-4; and American Journal of Sociology 64 (1959): 610-25. A
bibliography of other discussions of this problem appears in Otis Dudley Duncan et al..
Statistical Geography (Glencoe, 1961).

26a. If variances of the ten groups are similar.

27. Goodman, op. cit.

120

EXPLANATION

were concerned with race and literacy in 1840, he would not have been
able to run an individual correlation directly. The only data available
to him would have been aggregate census data. He would, therefore,
have to use regression techniques on his data to obtain an answer.

The ecological fallacy is not well named, for precisely the same
sort of error can be committed in the use of any kind of aggregate data.
The ecological fallacy could appear, for example, in the analysis of
generational as well as areal classes — in classes which are temporal
rather than spatial in nature. Historians, therefore, have still another
reason to be wary of this problem.

The fallacy of false extrapolation is a statistical series which
is stretched beyond the breaking point. It occurs in a variety of forms.
The most clear and simple is a generalization from a true series A, B
to a false A, B, C. The average size of the American family was re-
ported by the Census Bureau as 3.71 persons in 1940 and 3.54 per-
sons in 1950. To extrapolate from these known quantities, on the as-
sumption that the American family is shrinking at an arithmetical rate
of 0.22 persons per decade, to the conclusion that the American family,
in 2070, will consist of 0.90 persons, is absurd. The Census Bureau
estimated that family size actually increased from 3.54 to 3.65 in the
decade 1950-1960. To extrapolate from these figures on the assump-
tion that families are growing in geometrical ratio (the increment, let
us say, is doubling every decade) to the conclusion that the average
American family in 2070 will include 228.82 persons is not merely
absurd but inconceivable, in every sense of the word.

Many terrifying predictions of a great demographic Gotterdam-
merung lying just around the corner — not in 2070, but in 1970, or
1975, or 1980 — are not absurd, but they may be equally fallacious. There
is clear evidence that birth rates have been rising at a formidable rate. But
there has been a good deal of loose talk about population explosions that
fails to take into consideration the increase in contraception, the cumula-
tive effect of education and affluence upon birth rates, and other counter-
vailing phenomena. There is an ongoing crisis, but maybe not the
incipient catastrophe which some of our Cassandras so confidently ex-
pect.

There was an eighteenth-century American clergyman, Rev. Ezra
Stiles, who loved to weigh and measure things. He often talked of his
findings with his many visitors, and sometimes he even weighed and
measured them as well. But he outdid himself in 1760 with the follow-
ing projection of the denominational demography of New England:

FALLACIES OF GENERALIZATION 121
A.D. Episcopalians Friends Baptists Congregationalism

1760

12,600

16,000

22,000

440,000

1785

23,200

32,000

44,000

880,000

1810

46,400

64,000

88,000

1,760,000

1835

92,800

128,000

176,000

3,520,000

1860

185,600

256,000

352,000

7 MILLIONS'

* Quoted in Carl Bridenbaugh, Mitre and Sceptre (New York, 1962), p. 12.

Small wonder that this incorrigibly optimistic Congregational divine could
plan a monument for himself with a sketch of the universe on one side
and the Stiles family tree on the other! It is said that the only biblical
injunction which New England Congregationalists consistently obeyed
was the one which required them to increase and multiply — but not pre-
cisely according to Stiles's projections. 28

The field of demography is littered with discarded extrapolations
that seemed the soul of reason at the time they were made. The Negro
population of the United States appeared to fall sharply from 1860
to 1880, which led many intelligent Americans to predict the imminent
extinction of Negroes in this nation. A little later, others predicted from
an inverse correlation between IQs and fertility rates that the Republic
would soon be inhabited by a race of morons. Still others, from a differ-
ential in the reproductive rates of members of various religious groups,
predicted that the United States would become a fief of the Pope. A
satirical specimen of the fallacy of false extrapolation is supplied by
Mark Twain, who wrote:

The Mississippi between Cairo and New Orleans was twelve hundred and
fifteeen miles long one hundred and seventy-six years ago. It was eleven
hundred and eighty after the cut-off of 1722. It was one thousand and
forty after the American Bend cut-off. It has lost sixty-seven miles since.
Consequently, its length is only nine hundred and seventy-three miles at
present.

Now, if I wanted to be one of those ponderous scientific people, and
"let on" to prove what had occurred in the remote past by what had occurred
in late years, what an opportunity is here! Geology never had such a chance,
nor such exact data to argue from! Nor "development of species," either!
Glacial epochs are great things, but they are vague — vague. Please observe: In
the space of one hundred and seventy-six years the lower Mississippi has short-
ened itself two-hundred and forty-two miles. That is an average of a trifle over
one mile and a third per year. Therefore, any calm person, who is not blind
or idiotic, can see that in the Old Oolitic Silurian Period, just one million

28. The number of Congregational churches is thought to have increased from 465 in
1750 to 1,706 in 1850. But the number of Congregationalists seems to have increased
in a smaller ratio even than this. See Edwin S. Gaustad, Historical Atlas of Religion in
America (New York, 1962), pp. 62, 167, 168.

122

EXPLANATION

years ago next November, the lower Mississippi River was upward of one
million three hundred thousand miles long, and stuck out over the Gulf
of Mexico like a fishing-rod. And by the same token any person can see
that seven hundred and forty-two years from now the Lower Mississippi
will be only a mile and three-quarters long, and Cairo and New Orleans
will have joined their streets together, and be plodding comfortably along
under a single mayor and mutual board of aldermen. There is something
fascinating about science. One gets such wholesale returns of conjecture
out of such a trifling investment of fact. 29

Another kind of false extrapolation occurs in the form of a:b::A:B,
where a, b, and A are known, and an unknown B is falsely extrapolated.
An example may appear in a group of ingenious essays by two naval
historians, Michael Lewis and Lieutenant Commander D. W. Waters,
R.N., on the argument of the Spanish Armada. The authors were espe-
cially interested in the number of long-range guns (culverins) carried by
the great galleons of that mighty fleet. No direct evidence exists to answer
this question. Lewis and Waters were compelled to operate by means of an
extrapolation from three knowns to an unknown — from a known number
of Spanish galleons (a) carrying a known number of culverins (b) to
the unknown total number of culverins (B) carried by all the galleons in
the fleet (A ) . But galleons came in many assorted sizes, and culverins,
too. In the opinion of some learned students of this question, it is prob-
able that Lewis and Waters, for all their care and ingenuity, overestimated
the number of long guns in the Spanish fleet by this method. 30

£»> The fallacy of false interpolation is a form of overgeneraliza-
tion, in which a line is inaccurately run between two known points in
order to estimate the location of an unknown point in between. It is,
in other words, a false estimate of an unknown B from a known A and
C. There is no sure defense against this fallacy except full factual in-
formation. But there are two logical rules of thumb which, obvious
though they may be, are often overlooked.

First, the accuracy of interpolations of unstable indices must
necessarily vary inversely with the degree of their instability and with
the number of known index points. A grossly unstable index ought to be
interpolated with great caution and reluctance. Few indices are as un-
stable as the level of voter participation in a free and open polity. In
early American presidential years, participation fluctuated wildly from one

29. Mark Twain, Life on the Mississippi (New York, 1917), pp. 155-56.

30. Michael Lewis, "Armada Guns: A Comparative Study of English and Spanish Arma-
ments. Section V. The Guns of the Spanish Fleet, 1588," The Mariner's Mirror 29
(1943): 3-39; D. W. Waters, "The Elizabethan Navy and the Armada Campaign," ibid.,
35 (1949): 90-138; cf. Garrett Mattingly, The Armada (Boston, 1959), p. 416 n.

FALLACIES OF GENERALIZATION

123

election to another and from one state to the next. Moreover, the number
of known index points is small. In New Hampshire, the percentage
of adult males voting in three elections was:

If anybody tried to interpolate the turnout in 1824, solely from these
figures, he would be far off the mark. According to Richard McCormick,
who made the estimates above, only 16.8 percent of New Hampshire's
adult males came to the polls in 1824! 31

A second obvious rule of thumb is that an interpolation can be no
stronger than its weaker pole. A cautionary example is an article by
Professor George Rogers Taylor called "American Economic Growth
before 1840: An Exploratory Essay" in The Journal of Economic His-
tory 24 (1964) : 427-44.

Taylor's purpose was to develop a preliminary hypothesis, from
scattered statistical snippets and impressionistic material, as to per
capita growth rates in early America. He began with the work of an-
other economist who studied the same problem by working backward
from the period 1840-1959 where statistical evidence was comparatively
good, and the per capita growth rate seems to have been something
like l s /s percent per year. Could this rate apply to the period 1676-
1840? The other economist, Raymond Goldsmith, did not think so, for
it would require absurdly small levels of per capita income at the be-
ginning of the period. Instead, Goldsmith suggested an extended average
growth rate before 1840 of 0.6 percent per year, which never rose, in
his opinion, in any fifty-year part of this early period to as much as one
percent per year.

Taylor accepted much of this, but not the presumption of a steady
low growth rate. Instead he hypothesized from a variety of demographic
and economic materials that the per capita rate of growth in the period
1710-1775 was "relatively high for a preindustrial economy," perhaps
as high as one percent, maybe even higher. From this conclusion he in-
terpolated the additional conclusion that the average per capita growth
rate in the period 1775-1840 approached zero. "Per capita income
in 1840 was about the same or at least not substantially higher than in
the early 1770's," he guessed.

This interpolation is logically tenable. If Taylor's estimates of

31. Richard P. McCormick, "New Perspectives on Jacksonian Politics," American
Historical Review 65 (1960): 288-301.

1808

1812
1828

62.1%
75.4%
76.4%

124

EXPLANATION

growth rates before 1775 and after 1840 are correct, then his interpo-
lation of a zero-sum in the intervening period must be right, too. But
if one of the guesses is wrong, then the interpolation collapses. Taylor's
figures for the post- 1840 period seem acceptable. But his estimates for
the period 1710-1775 are derived from partial and contradictory
evidence. A more open hypothesis would, in my judgment, be more
serviceable to sustained research. 32

The fallacy of the insidious generalization is committed by a
historian who swears up and down that he cannot and will not generalize
upon his subject, and then proceeds to bootleg generalizations into his
work, without recognizing their existence or controlling their content.
An example is H. A. L. Fisher, an English academic historian who
prefaced one of his books with the following assertion: "One intellectual
excitement has, however, been denied me. Men wiser and more
learned than I have discerned in history a plot, a rhythm, a predetermined
pattern. These harmonies are concealed from me. I can see only . . .
one great fact with respect to which, since it is unique, there can be no
generalizations." 33

This statement is followed by a considerable number of generaliza-
tions, some of which are quite as absurd as the disclaimer which preceded
them. 'The Athenian empire, the brilliant growth of two generations,
shared the fate of every polity which rises by the repression of local
liberties," Fisher wrote. Again: "It is the property of polytheism to be
tolerant." And again: "Men once embarked on the ocean of political
strife are apt to be carried further than they intended." 34

One might note, in Fisher's prefatorial statement, a common con-
fusion in the author's mind between generalizations concerning the whole
of history and generalizations of a limited sort within history. A good
many historians, I think, have condemned both of these things together.

The fallacy of the insidious generalization commonly occurs in a
form even more insidious than Fisher's errors. Impressionistic historians,
who bitterly inveigh against quantification and its dehumanizing tend-
encies, are quick to use the words "few," "some," "most," "many,"
"singular," "typical," "exceptional," "common," "customary," "normal,"
"regular," "recurrent," "periodic," "widespread," "often," and many
others. These terms imply numbers, and. numbers need counting. And

32. For a historiographical model of interpolation by a series of orderly inferences, see
Willi Apel, Gregorian Chant (Bloomington, Ind., 1966), pp. 33, 507-15.

33. H. A. L. Fisher, A History of Europe, 3 vols. (London, 1935), l:vii.

34. Ibid., 1:32, 45, 67.

FALLACIES OF GENERALIZATION

125

yet they are used without quantification, by scholars who believe that
quantifying is no part of their job. The inaccuracy of much impres-
sionistic history is an inevitable result. Other historians make statements
in which even these words are implicit. The results are even worse. So
common is this form of error that the reader can pick up almost any
work of history from his shelves, turn to almost any page, and find an
example.

$*» The fallacy of the double-reversing generalization is the op-
posite of many errors in this chapter — a halting generalization rather
than a hasty one, an understatement rather than an overstatement. It is
a species of interpretative bet-hedging, which in an extreme form be-
comes no interpretation at all but a maze of mutual qualifications or a
cunning balance of casuistical contradictions, or a trackless wilderness
of pettifogging detail, or a slippery ooze of substantive (as well as
semantical) shilly-shally.

A familiar example is Theodore Roosevelt's first presidential state
paper, in which the author, with uncharacteristic restraint, tried to play
things both ways on a number of controversial issues of the day. The
most famous part of this document concerned public regulation of
corporate enterprise — a passage which the inimitable Mr. Dooley took
great pleasure in satirizing: "Th' trusts," says he, "are heejoous
monsthers built up be th' enlightened intherprise iv th' men that have
done so much to advance progress in our beloved country," he says.
"On wan hand I wud stamp thim undher fut; on th' other hand not so
fast."

There are many comparable examples in historical scholarship.
More than a few monographers have timidly advanced a tentative thesis
in Chapter 1, only to withdraw it in Chapter 2 with an "on th' other
hand not so fast." An example is Alfred F. Young's The Democratic
Republicans of New York: The Origins, 1763-1797 (Chapel Hill, N.C.,
1967). This distinguished work possesses many great merits in the
breadth and depth of its accurate and painstaking research. But it is more
satisfactory in its parts than in the whole. The author's idea of truth,
like Burke's idea of liberty, seems to be "a sort of middle" between in-
terpretative extremes. There is, of course, no necessary fallacy in
moderation. But Young's assertions are so thoroughly amended by strings
of qualifications that the main point is neutralized, or obscured. And
sometimes he declares unequivocally that X was the case, only to de-
clare unequivocally a little later that not- A' was the case.

What, for instance, was George Clinton's conception of the economic

126

EXPLANATION

role of the state government? On page 56 Young characterizes it as
"essentially negative." On page 570 he calls it essentially "positive" on
the same questions.

What was the nature of the relationship between Federalist leaders
and their constituents? We are informed on one page that "most Fed-
eralist support was uncoerced," but also that "to a considerable extent
Federalist electoral strength continued to be a vote of economic de-
pendents" (p. 569). A reconciliation is conceivable between these two
statements, but none is supplied by the author.

What, precisely, was the degree of continuity and discontinuity be-
tween the political groups of 1788 and the parties of the 1790s? The
author seems to stress elements of continuity, but there is no clear and
conclusive estimate of this difficult problem. Instead the reader learns
that "In the leadership of the Federalists, the continuities in New York
politics from the Revolution through the 1 790's were striking. . . . There
were also discontinuities. . . . The leadership of the Republicans showed
similar continuities and discontinuities . . ." (pp. 566-67).

The ambivalence of Young's interpretation of the impact of the
French Revolution upon New York politics is captured in the tension
between the beginning and the end of his essay on that question. The
chapter is called "The Spirit of 1776 Rekindling." But the author
concludes "it was not quite true of New York, as Jefferson said of the
country as a whole, that 'all the old spirit of 1776 is rekindling.' " Never-
theless, he insists "a fire had indeed been lit," and this reader, at least,
lost his way in billowing clouds of conceptual smoke (pp. 345-65 ).

A good deal of the confusion in Young's book is a consequence of
impressionistic research and of his attempt to compress an analytical
thesis into a narrative form; but much, I think, is substantive as well.
The complaint is not that he took a middling position, but rather that
his position tends to shift within an unstable range of middling possi-
bilities.

Another example is Norman K. Risjord's The Old Republicans:
Southern Conservatism in the Age of Jefferson (New York, 1965). A
fair judgment upon this work is passed by Noble Cunningham:

The author begins by picturing the old Republicans in traditional shades
as a group who appeared in reaction to the surge of nationalism following
the War of 1812, men who took to heart the compact theory of government
and adhered to the "principles of 1798." Although this theme is later
restated, it is so modified in other places as to raise the question of how well
the author has thought through his interpretations. While stating as fact
that the old Republicans appeared as a reaction to postwar nationalism,
the author later writes that this "is generally assumed" but "only partially
true." . . . Throughout the work, shifting emphasis and conflicting interpre-

FALLACIES OF GENERALIZATION 111

tations leave no clear picture of the old Republicans or of Southern con-
servatism. 55

The fallacy of the overwhelming exception is a generalization
which is a good deal less general than it appears to be. The authors of
1066 and All That assert, for instance, that "Magna Charta was there-
fore the chief cause of Democracy in England, and thus a Good Thing
for everyone (except the Common People)." 36

A serious example appears in an essay by an American business
historian, Gordon C. Bjork, who argued the thesis that "the relatively
slow rate of population growth in the [American] seaport cities under
consideration for the period 1790 to 1825 must be blamed, in large
part, on the failure of foreign trade to expand." He contended that some-
thing called "trade gain per capita" fell from a base of 100 in 1790
to 30 in 1825 — in Boston, New York, Philadelphia, and Baltimore.

Bjork's critical index of "trade gain per capita" is a complicated
thing, calculated from three other indices, all of which are incorrect.
But our business is merely with his estimates of the expansion of com-
merce in the cities. We discover that he omitted cotton from his calcula-
tions, though it was the most expansive domestic commodity of the
period and important even in the trade of ports as far north as Boston.
Moreover, he excluded re-exports, which other scholars have found to
account for much of the expansion of American commerce during the
period 1790-1807. Bjork justified his exclusions on the ground that
cotton and re-exports were not produced in the western hinterlands of
the seaport cities — which is irrelevant to any estimate of trade volume.
All four of these cities had southern hinterlands as well as western ones;
all of them were engaged both in the re-export business and in the
transshipment of cotton.

These exclusions are not readily apparent to a casual reader of
Bjork's paper. But they are vital to his interpretation. To make such
sweeping exclusions is to make nonsense of the subject. 37

35. American Historical Review 71 (1966): 1062.

36. Walter C. Sellor and Robert J. Yeatman, 7066 and All Thai, new ed. (London,
1965), p. 34.

37. Gordon C. Bjork, "Foreign Trade," in David Gilchrist, ed.. The Growth of the Sea-
port Cities (Charlottesville, 1967), pp. 54-61. The accuracy of Bjork's interpretation is
diminished by other errors. A weighted average of the four seaport cities would show
a different result from the one which he calculated. Moreover, Bjork appears to have
inflated his population index, and his calculation of terms of trade contradicts, without
explanation, the source which he cites: Douglass North's The Economic Growth of the
United States, 1790-1860 (Englewood Cliffs, N.J., 1961).

128 EXPLANATION

$»» Positive rules for the construction of sound statistical generali-
zation need not be discussed here. They are thoroughly explored in statis-
tics texts. But there is something of another sort which needs to be said.
The structure and function of generalization in history has been con-
siderably confused by the errors of a reigning school of epistemologists,
who hold that every explanation in history (and indeed everywhere
else) must consist in referring the thing to be explained to a "general
law" or "universal hypothesis" or "hypothesis of universal form." These
three phrases, suitably shrouded in shudder-quotes, come from the classic
argument for this interpretation by Carl G. Hempel. His thesis is:

The explanation of the occurrence of an event of some specific kind E at
a certain place and time consists, as it is usually expressed, in indicating the
causes or determining factors of E. Now the assertion that a set of events —
say, of the kinds Ci, C2 . . . , Cn — have caused the event to be explained,
amounts to the statement that, according to certain general laws, a set of
events of the kinds mentioned is regularly accompanied by an event of
kind E. Thus, the scientific explanation of the event in question consists of

( 1 ) a set of statements asserting the occurrence of certain events C v . . . C„
at certain times and places.

(2) a set of universal hypotheses, such that

(a) the statements of both groups are reasonably well-confirmed by
empirical evidence,

(b) from the two groups of statements the sentence asserting the occur-
rence of event E can be logically deduced.

Hempel's criteria for a "general law" or "universal hypothesis" are
rigorous. A law must be universally true in all cases where a certain
complex of conditions is satisfied. And an explanation model must be
"logically conclusive," in the literal sense that the conclusion must follow
necessarily from the premises. The only complete explanation is an
explanation which is "deduced by logical reasoning" in this way. An
explanation is not complete unless "it might as well have functioned as a
prediction." 38

38. Hempel, "The Function of General Laws in History," as reprinted in Patrick Gar-
diner, ed., Theories of History (Glencoe, 111., 1959), pp. 344-56. Hempel's classic article,
published in 1942, was not, of course, the first statement of the "Hempelian model."
The idea that explanation consists in showing that something obeys a law has been
traced back to Berkeley, and probably appeared much earlier. (E. Meyerson, De
I' Explication dans les Sciences, 2d ed. [Paris, 1927]). More recently it appeared in forms
very close to Hempel's in Karl Popper's Logik der Forschung (Vienna [19341), and in
Morris Raphael Cohen, Reason and Nature (New York, 1931). It is called by many
different names — the "regularity interpretation" (Patrick Gardiner, The Nature of
Historical Explanation, p. 65); the "covering law model" (Willian Dray, Laws and Ex-
planation in History, p. 1); and the "deductive model" (Michael Scriven, "Truisms
as Grounds for Historical Explanations," Theories of History, p. 444.

FALLACIES OF GENERALIZATION

129

But these conditions cannot be met in historical writing. Consider a
whimsical example of a universal law which is presented by another
philosopher, Morton White. "Let us assume that the statement 'When-
ever Chinese eat, they eat with chopsticks' is true," he writes. But this
statement, as it stands, is of course untrue, in a way which invalidates
other alleged universal laws in history. First, Chinese have not always
eaten with chopsticks. The earliest humanoid inhabitants of the Yellow
River valley probably lapped up their dinners like dogs. Moreover,
today's Chinese do not eat with chopsticks everywhere. It is unlikely
that a well-mannered Mandarin, however loyal he may be to the ways
of his ancestors, would eat with chopsticks at Maxim's. Most important,
even when the use of chopsticks was at its peak in China, most Chinese
did not use them some of the time (i.e., mourning periods), and some
Chinese did not use them most of the time (i.e., fingerless Chinese).
One must, therefore recast White's statement to read, "Within certain
limits of time and space, Chinese tended to eat with chopsticks."

This statement differs from a universal law in two respects. First,
it applies differentially through time within a closed temporal frame of
reference. Second, it is a statistical statement. An explanation model
which includes a temporalized statistical generalization statement of this
sort cannot support the sort of deduction which the Hempelian model
requires. A probability statement asserts that C is sometimes followed by
E, and sometimes not. Maybe E happens 999 times out of 1 ,000, but if
it doesn't happen once there can be no necessary and certain deduction

In revisions of his thesis, Hempel has stressed a second kind of covering law model,
in which the explanans does not include a universal law, but rather a statement of
statistical probability. This second model is more relevant to the explanations which
historians (and natural scientists) actually make. And it is radically different, in its
logical behavior, from the deductive model of explanation. But even with this salutary
modification, serious problems remain. Hempel still seems to insist that all explanations
must conform to one or the other of his models. (He would allow "explanation sketches"
which are understood as incomplete statistical statements to be filled in by some future
empricist.) Now, if it is true that no historian (or natural scientist) can hope to find a
universal law of the sort required by Hempel's first model, then it follows that all
explanations in history (and in the natural sciences) must be statistical or proto-
statistical in nature. And this, I think, is the quantitative fallacy, which returns to the
Pythagorean error of thinking that all significant things are numbers.

Moreover, Hempel seems to argue that only statements of very high ("close to 1")
probability can be used in explanations. This is surely unsatisfactory. How close to 1?
What criteria can there be? And what, precisely, is the logical relationship between
an explanandum and an explanans which contains a statistical statement? Is it dif-
ferent from an explanation model which contains a narrative statement, or a moti-
vational statement, or a group composition statement? I don't see how. In each of these
types, the bond between an explanans and an explanandum must be presently under-
stood in psychological and/or pragmatical terms, perhaps because it conforms to a
logic we don't yet understand.

130

EXPLANATION

from the premises to the conclusion. This is true, I think, equally in the
natural sciences, where universal laws are equally elusive. Albert Einstein
wisely asserted that "As far as the laws of mathematics refer to reality
they are not certain, and as far as they are certain they do not refer
to reality." 39 The same thing might be said of Hempel's general laws.

Philosophers of history have recently squandered a good deal of
time and effort in a vain attempt to adjust the Hempelian model to the
plain fact that historians do not use universal laws in their work. A
few have attempted to adjust the plain fact to the Hempelian model.
Some extraordinarily ingenious arguments have been invented, but the
enterprise is, at bottom, absurd. Analytical philosophers have failed,
I think, to be sufficiently empirical in their own work. They have failed
to give sufficient attention to historical thought as it actually happens.
And they have erected a deductive standard which is so exalted that
no mortal empiricist can ever begin to meet its requirements. One
wishes that analytical philosophers might harken to the words of
Robert Frost:

As ever when philosophers are met,

No matter where they stoutly mean to get,

Nor what particulars they reason from,

They are philosophers, and from old habit

They end up in the universal whole

As unoriginal as any rabbit. 40

39. James R. Newman, ed., The World of Mathematics, 4 vols. (New York, 1956),
3:1646.

40. Robert Frost, "The Lesson for Today," Complete Poems (New York, 1965), p. 474.
To this unkindness, analytical philosophers could conceivably reply with a poem which
is not by Robert Frost. It is called "The Lesson for Yesterday."

Historians, rather like primitive moles,

Live purposeless lives in particular holes,

Which they dig with their noses, or else with their toeses

(A few have invented small shovels and hoeses).

They're burrowing blindly in Byzantine tunnels

Constructed like sinuous serpentine funnels;

They're burrowing busily, back to the past:

A steady regression to nowhere — fast.

CHAPTER V

FALLACIES OF
NARRATION

Historians begin by looking backward. They often end
by thinking backward.

— Nietzsche

Most historians tell stories in their work. Good historians tell true
stories. Great historians, from time to time, tell the best true stories which
their topics and problems permit. Narration is not the only form of
explanation they use, but it is one of the more common and most
characteristically historical forms. 1

A story explains how and what — not why. For many epistemologists
this is no explanation at all. But if explaining is understood to mean
what every dictionary says it means, and what common usage makes it
mean — namely, making clear, plain, and understandable — then story-
telling is a form of explanation which is common not merely in his-
torical scholarship but in a dozen other disciplines and in daily life as
well. When a policeman arrives at the scene of an accident and asks,
"What happened?" he wishes to hear a true story. When a child asks,
"What was World War II?" or "Who was Winston Churchill?" he wants

1. This point, important as it is, has recently been exaggerated by epistemologists. Note
three errors in the following statement: "The difference between history and science is
not that history does and science does not employ organizing schemes which go be-
yond what is given. Both do. The difference has to do with the kind of organizing
schemes employed by each. History tells true stories." (Arthur C. Danto, Analytical
Philosophy of History [Cambridge, 1965], p. 111).

First, history tells no tales — historians tell them. Second, scientists are storytellers
too. All scientists use story schemas some of the time; some use them nearly all the
time. Examples are geologists, paleontologists, and astronomers. Third, historians use
many other organizing schemes in their work, occasionally in place of narratives, and
often in addition to them. It is, I think, an error to distinguish history from the sciences
on epistemological grounds alone.

132

EXPLANATION

and usually gets an explanation in the form of a short narrative history of
the war or a simple narrative biography of the statesman.

Sometimes narrative explanations are very short and simple and yet
sufficient to the task at hand. The question "What happened at the
battle of Gettysburg?" might be answered by the shortest story imagi-
nable — Lee lost. No generalizations are required, and no cause, no motive,
no elaborate composite-models, no analogies, and no pedantical detail.

But in anything more than this atomic form, storytelling becomes
exceedingly complex. This chapter will be confined to merely one elemen-
tary aspect of an immensely intricate subject — the problem of time and
temporal integrity in a narrative. A sense of time is not a simple thing.
Time is something which people must painfully learn to think clearly
about — something, indeed, which they must be taught. And there are
many obstacles in the path of understanding it. It is difficult enough in
this day and age for a person to imagine that there really was a past
and that there will be a future. But even when that lesson is learned,
one must master the idea that there are many different pasts and futures,
and many different degrees of pastness and futurity. A child's history of
the United States is likely to locate all past events on a single plane.
Winthrop, Washington, Lincoln, and Kennedy are dimly perceived as
chronological contemporaries, without depth, breadth, or temporal
discrimination.

Psychologists have demonstrated that these discriminations are
difficult for a young mind to master. It is hard for children to understand
that Philadelphia is simultaneously north of Baltimore and south of New
York — particularly if they happen to live in Boston. In precisely the
same way, they cannot quickly grasp the idea that Washington was
simultaneously in Winthrop's future and Lincoln's past. Adults often
have this trouble, too. The dimness of their temporal perceptions may
be the deeper reason for many so-called factual mistakes. It was said
of Mrs. Disraeli, for example, that "she could never remember who came
first, the Greeks or the Romans." 2 Mrs. Disraeli's difficulty may have de-
rived from the fact that Greeks and Romans really existed for her as
chronological contemporaries.

The memories of many professional historians seem to work a little
better than Mrs. Disraeli's. But not much better. Some symptoms of time
trouble are the following fallacies.

The fallacy of anachronism is the common denominator of
most forms of error in this chapter. It generally consists in the description,

2. Robert Blake, Disraeli (New York, 1968), p. 146.

FALLACIES OF NARRATION

133

analysis, or judgment of an event as if it occurred at some point in time
other than when it actually happened. If the event is located too early
in time, then the error is sometimes pedantically called a "prochronism."
If it is made to happen too late, then the result is a "metachronism."

Anachronisms nourish in a profusion of particular forms. A simple
mistaken date is technically an anachronism, as when a recent student
of Benjamin Franklin somehow made his hero die in the wrong year. 3
So also is a single misplaced object, event, or word. The title of William
Allen White's biography of Calvin Coolidge provides two examples in
four words — A Puritan in Babylon.*

These howlers are unimportant in themselves, except as signs of
more serious analytical anachronisms, by means of which many historical
problems have been grossly misconceived. Consider the case of an
iconoclastic book called Jefferson and Civil Liberties: The Darker Side
(Cambridge, 1963), in which the author, Leonard Levy, committed
an analytical anachronism of serious proportions.

Levy argued that there was "a strong pattern of unlibertarian or
even antilibertarian thought and behavior extending throughout Jeffer-
son's long career," and specifically that Jefferson countenanced loyalty
oaths, internment camps, bills of attainder, Blackstonian prosecutions
for seditious libel, censorship of books and words and even thoughts,
infringements of academic freedom, and violations of various provisions
of the Bill of Rights. This indictment is accompanied by epithets such
as "obnoxious," "abhorrent," "sophistical," "narrow and ritualistic,"
and many disparaging comments upon Jefferson's mind and character. 5

Critical reviewers have brought some of Levy's factual statements
into dispute. 6 But there is a deeper deficiency in the book. Levy asserts
that "during Jefferson's lifetime, there was never an issue for which
incontestably familiar libertarian standards were lacking to guide his
judgment. Experience with the application of certain of those stan-
dards may have been slim, yet the standards themselves had been estab-
lished." 7

This, I believe, is a very great mistake on Levy's part. It is factually

3. John H. Best, Benjamin Franklin on Education (New York, 1962), p. 7.

4. (New York, 1938); see also H. P. Gambrell, Mirabeau Buonaparte Lamar: Trou-
badour and Crusader (Dallas, 1934); Carleton Mabie, American Leonardo: A Life
of Samuel F. B. Morse (New York, 1943); and S. H. Brockunier, The Irrepressible
Democrat: Roger Williams (New York, 1940).

5. See, e.g., pp. xii, 27, 40, 42, 50, 55. An exception is the first chapter, in which Jeffer-
son is praised for his religious libertarianism.

6. A good and fair review by Joseph H. Harrison, Jr., appears in The William and Mary
Quarterly, 3d ser. 21 (1964): 451-54.

7. P. 20.

134

EXPLANATION

false to argue that there were incontestably familiar libertarian standards
in Jefferson's own day. Levy himself, and others, have demonstrated
that there were many different ideas of liberty which were eminently
contestable — and constantly contested in that era. There were new ideas
of free expression, which some of Jefferson's younger supporters avowed
but which Jefferson did not. And there were other conflicting ideas
in various stages of development, with a good deal of confusion in
between.

Moreover, there were many difficult problems in the early Republic
for which no "incontestably familiar libertarian standards" exist, even
today: the problem of civil liberties in wartime, the problem of the ex-
tension of liberty to people who are determined to destroy the liberty
of others — these are questions for which there are no easy answers.
Jeffersonian answers were sometimes of the sort which Levy and the
American Civil Liberties Union condemn as "unlibertarian." But always
Jefferson's responses were of such a nature that Jefferson himself could
honestly and consistently reasonably claim them as libertarian — accord-
ing to his lights.

Levy formed in his own mind an idea of what civil liberties should
entail — an idea which has some relevance in some of its particulars to
some of Jefferson's associates (men younger than Jefferson himself).
Then he proceeded to condemn Jefferson, sometimes explicitly, some-
times by innuendo, for not living up to this exalted atemporal standard.
In short, Levy analyzed and evaluated Jefferson by measuring his acts
and attitudes against the standards of the ACLU and tallying all the
discrepancies. The result is objectionable not merely because it is unf au-
to Jefferson but also because it distorts and falsifies the texture of Jeffer-
sonian thought.

Anachronisms of this sort are sometimes so subtle and intricate
that they consistently elude analysis and understanding. A disturbing
case in point is the work of Douglas Southall Freeman, who possessed
an extraordinary talent for seven- volume studies of Victorian gentle-
men — which Robert E. Lee was, but George Washington emphatically
wasn't. Nevertheless, Freeman clapped a cocked hat on the head of
General Lee, and called him General Washington. The underlying
errors are exceedingly difficult to locate with precision. But somehow
the psychic processes are wrong — the attitudes and responses of Free-
man's Washington belong to a man born in 1807 rather than in 1732. 8

Complex anachronisms are not merely matters of sterile academic

8. Cf. the criticism of Perry Miller in The William and Mary Quarterly, 3d ser. 9
(1952): 226-29.

FALLACIES OF NARRATION

135

interest. They are snares that have caught up many people in the world.
Consider the case of the Cox Commission report, an attempt by a group
composed mostly of lawyers to produce something like a history of the
student rebellion at Columbia University in 1968. The authors were able
and intelligent men, but their honest attempt to understand the thought
patterns of another generation was marred by analytical anachronism.
They sought to explain the discontent of student rebels by quoting a
character in J. D. Salinger's Franny and Zooey\ My students laughed
uproariously at this gaffe. They suggested that the Commission might as
well have studied the works of Scott Fitzgerald."

The fallacy of presentism is a complex anachronism, in which
the antecedent in a narrative series is falsified by being defined or in-
terpreted in terms of the consequent. Sometimes called the fallacy of
nunc pro tunc, it is the mistaken idea that the proper way to do history
is to prune away the dead branches of the past, and to preserve the
green buds and twigs which have grown into the dark forest of our con-
temporary world. Imagine two historians, both studying the same prob-
lem, which happened thus:

(Present)

Time

Events

1

A

B C

D

E

2

B

C D

E

F

3

C

D E

F

G

4

D

E F

G

H

t 5

E

F G

H

I

A total of twenty-five events actually happened, in this hypothetical
matrix of pastness. 10 Suppose that each historian can select fifteen of
them. The presentist chooses the following:

9. Archibald Cox et al., Crisis at Columbia: Report of the Fact-Finding Commission
Appointed to Investigate the Disturbances at Columbia University in April and May,
1968 (New York, 1968), p. 23.

10. This matrix, like those which follow, is meant merely as an heuristic device. There
is, of course, no such symmetry in any past problem. A more realistic representation
might be provided by the following method. Take a Jackson Pollock painting and cut
it into a jigsaw puzzle with a hundred thousand parts. Throw away all the corner pieces,
two-thirds of the edge pieces, and one-half of the rest. What remains is a little closer
to the problems historians actually study.

136

EXPLANATION

Time Events

(Past) 1 E

2 E F

3 E F G

4 E F G H
(Present) 5 E F G H I

The other historian, who is not a presentist, selects the same number of
events by a different method of sampling. The result is:

Time Events

(Past) 1 A B C D E
2

3 C D E F G
4

(Present) 5 E F G H I

The latter pattern not merely provides a more satisfactory sense of the
configuration of past events, but also a more enlightening perspective
upon the present. The presentist method is self-contradictory. In the
name of modernity and relevance and utility, it sacrifices precisely that
kind of knowledge which historians can most usefully provide: knowl-
edge useful in the establishment of present trends and future tendencies.

There are many examples in historical scholarship. One of them
is a standard textbook in intellectual history, John Herman Randall's
The Making of the Modern Mind, first published in 1 926. The author's
avowed purpose was "to explore the mind of the present generation,
to unravel the many threads that enter into its tangled fabric and trace
them back to their first appearance in the loom of history." 11 The result
is the usual humbug about Hebrews and Greeks, the "modernity" of the
Renaissance and the "democracy" of the Reformation. 12 Each of
Randall's backward projections of present phenomena so grossly dis-
torts the past that the reader receives an utterly erroneous idea of
events in earlier periods, and of tendencies in his own as well. A reader
who wishes to discover the damage which is done by Randall's un-
fortunate method might examine his version of Copernicus, whom he
makes into a John-the-Baptist for Isaac Newton, and compare it with a
contrasting interpretation by an excellent historian of science, Thomas

11. John Herman Randall, The Making of the Modern Mind, rev. ed. (Boston, 1940),
p. 4.

12. Ibid., pp. 144, 165.

FALLACIES OF NARRATION

137

S. Kuhn. 13 In terms of our matrices, Randall interpreted Copernicus
in period one as E; Kuhn represented him as ABCDE. The two results
are radically different, and Kuhn's is surely superior.

The fallacy of presentism is a common failing in historical writing
by men who have never been trained in the discipline of history. A very
great man, in his own held, may serve as a second example. Bernard
Berenson was an art critic of splendid gifts. But he wasn't a historian
— which was unlucky for him, and unhappy in its consequences, for he
wrote several fat volumes of what are commonly called "art history."
In a lively little volume called Aesthetics and History he included a
chapter on "significant events in history," in which he made explicit
his criteria of selection. Berenson's view was precisely the same as
Randall's.

Significant events [he wrote] are those events that have contributed to making
us what we are today ... art history must avoid what has not contributed to
the main stream, no matter how interesting, how magnificent in itself. It
should exclude, for instance, most German and even Spanish and Dutch
art. It should dwell less and less on Italian art after Caravaggio, and end
altogether by the middle of the eighteenth century with Solimena and
Tiepolo. Except for Ribera, Murillo, Velasquez and Goya in Spain, and
Schongauer and Diirer and Holbein in German lands, the painters of
these countries are neither in the main line of development nor of universal
appeal to cultivated Europeans. 14

Berenson was a historical relativist when he wrote this book. He
allowed that every culture must have its own art history. But for himself
and his culture, he casually dismissed as nonevents all artistic accom-
plishments "from western Kamchatka to Singapore, from Greenland's
icy mountains to Patagonia's stormy capes, in Africa and on the islands
of the sea." And he was equally cavalier with all "the arts of China and
of India, remarkable and deeply human as they are." 13 They might
do for a Chinaman or an Indian, he conceded, but they "are not history
for us Europeans," he wrote.

Berenson's outlook, I think, differs from that of many other art
historians only in the degree to which its assumptions are spelled out,
and in the zeal with which they are extended to their logical conclusion.
Other art historians are more moderate in their statements, if not in
their thoughts. But a glance at the syllabus of an art history survey
course or the contents of a text in this field suggests that more than a

13. Thomas S. Kuhn, The Copernican Revolution (Cambridge, Mass., 1957); The
Structure of Scientific Revolutions (Chicago, 1962).

14. Bernard Berenson, Aesthetics and History (New York, 1948; 2d ed., 1954), pp.
257-58.

15. Ibid., p. 257.

138

EXPLANATION

few art historians are still chugging along Berenson's main line, pulling
passenger cars full of culture-hungry coeds. The main line isn't quite
as narrow-gauged as it used to be, there are more side excursions, and
the locomotives are a little more powerful. But the trip itself is much
the same.

Academic historians are not exempt from the same error. Consider
the case of Geoffrey Barraclough, a distinguished medievalist, now
converted to "contemporary history." Barraclough has published two
enthusiastically present-minded volumes, premised upon the notion
that since 1943, or thereabouts, traditional European historiography
has been reduced to irrelevance. "The traditional Europe — the Europe
of our history books, the Europe of Louis XIV and Napoleon and
Bismarck — is dead and beyond resurrection, and we may disabuse our
minds of the illusion that there is any special relevance, from the point
of view of contemporary affairs, in studying those neolithic figures,"
he believes. 16

I can't imagine that Barraclough means to be taken literally, but
he clearly intends to be taken seriously.

It [is] only a pardonable exaggeration [he writes], to say that, for me, it was
the Russian victory at Stalingrad in 1943 that made a total revision of
European history imperative. I realized with consternation that three years
in an English and two years in a German university had left me for all
practical purposes ignorant of eastern European history, save at a few
points where it caught the limelight. ... I knew a great deal of the
machinery of the papal chancery in the thirteenth and fourteenth centuries,
and of writings of the canonists; but I knew nothing of the Piasts, Przemy-
slids and the Ruriks. ... I had some notion of the place of the emperor
Charles IV in history, but knew little more than the names of his con-
temporaries, Louis the Great of Hungary and Casmir the Great of Poland.
I had read all the surviving letters of St. Boniface, the apostle of the Germans,
but nothing of St. Cyril and St. Methodius, the apostles of the Slavs. 17

It is impossible not to sympathize with Barraclough in his protest
against the narrow parochialism of so many (though not all) European
historians during the past two centuries. But his method would mark the
end not merely of Eurocentric history but of history itself. What, one
wonders, would be the task of ancient and medieval historians if the Ger-
mans had won the battle of Stalingrad? By Barracloughian logic, Peter the
Great, Alexander I, and even the great Lenin himself would then become
unpersons, instead of Bismarck & Co.; and our survey courses would serve
up a spicy schnitzel of Aravisci, Vangiones, and Triboci, in place of
a goulash of Piasts, Przemyslids, etc.

16. Geoffrey Barraclough, History in a Changing World (Oxford, 1955), p. 217.

17. Ibid., p. 10.

FALLACIES OF NARRATION

139

Surely any standard of significance which requires a 180-degree turn
in interpretations of the history of Europe before the fourteenth century,
because of the outcome of a battle in 1943, is not merely mistaken but
absurd. 18

No discussion of presentism in history can be complete without the
classic example of the "Whig interpretation of history," which has been
defined by Herbert Butterfield as the "tendency in many historians to write
on the side of Protestants and Whigs, to praise revolutions provided they
have been successful, to emphasise certain principles of progress in the
past and to produce a story which is the ratification if not the glorification
of the present." 19

I have heard historians use the pejorative phrase "Whig history" as
a synonym for any kind of presentism, which is unfair to Whiggery. The
same sort of error appears in works by scholars of all political persuasions.
There is, for instance, an anti-Whig history which commits the same fal-
lacy in an inverted form. Examples include the later works of Henry
Cabot Lodge and Henry and Brooks Adams. The latter, it is said, used to
greet each day by singing a song of his own invention, which consisted
entirely of three repeated words: "God damn it! God damn it! God damn
it!" For these gentlemen, history was indeed one goddamned thing after
another — a steady spiral running downward toward the left, and cul-
minating in some dark catastrophe — lava flowing through the streets
of Quincy, or a tidal wave crashing upon Nahant, or a wild-eyed mob of
Jews and Irishmen smashing in the doors of the Boston Athenaeum
and scribbling madly in the margins of books.

Presentism equally appears in the new-liberal narratives of Arthur
Schlesinger, Jr., where American history is the steady progress of prag-
matic liberalism from Jefferson to Jackson to Franklin Roosevelt. Finally,
the Kennedys become Top Family, and History comes to a .

The so-called New Political History does not display this deficiency
in so gross a form. As historians begin to fill their monographs with
more sophisticated analyses, the temptations for this species of simplicity
tend to diminish. But presentism still appears, in some surprising forms.
A case in point is William Nisbet Chambers' Political Parties in a New
Nation, an analytical history of the development of party structure in
the United States. For Chambers, parties are a Good Thing. "If party

18. Pieter Geyl, "Geoffrey Barraclough, or the Scrapping of History," Encounters in
History (Cleveland, 1961), pp. 336-40.

19. Herbert Butterfield, The Whig Interpretation of History (London, 1931), Preface.
Butterfield took a different position in a subsequent work, The Englishman and His
History (Cambridge, 1944), thus permitting E. H. Carr to have a little fun with "the
proto-Butterfield" and "the deutero-Butterfield," in What Is History? (New York,
1962), p. 51.

140

EXPLANATION

politics are contrasted with the near-chaos of faction politics that preceded
the formation of parties," he wrote, "it becomes apparent that even a
loose party system brings substantial advantages in both democratic
responsibility and coherence." 20

Chambers understands faction politics as pre-party politics. He
defines the antecedent in terms of the consequent, and the result is a
falsification of the former. Faction politics, of course, were not at all a
condition of near-chaos, before Jefferson and John Beckley brought order
and light. Factions functioned very well indeed in a different kind of
political environment. As that environment changed, new party forms
were created to serve new functions. Chambers' retrospective symmetry,
in short, distorts both the factions which preceded parties and the parties
themselves.

The problem of retrospective symmetry in historical narration is a
tough one, for without retrospection there can be no history, and without
symmetry there can be no narrative. But whenever retrospective symmetry
takes such a form that parts of the story are falsified, then the fallacy
of presentism results.

$»» The converse of presentism is the antiquarian fallacy, for
which a brief note should suffice. An antiquarian is a collector of dead
facts, which he stuffs full of sawdust and separately encloses in small
glass cases. Often, he is a gentleman (or lady) of respectable origins
who is utterly alienated from the present. The past serves him as a
sanctuary from a sordid world which he neither accepts nor understands.
His matrix of pastness looks a little like this:

Time Past events

(Past) 1 A B C D

2 BCD

3 CD

4 D

(Present) 5 $(§#%?!*<#**!!

At the Boston Athenaeum, one may discover flocks of tiny birdlike
old gentlemen, who nest in eery piles of dirty yellow paper and brood
their myths and memories into monumental Lives-and-Letters. In every

20. William Nisbet Chambers, Political Parlies in a New Nation (New York, 1963),
Foreword.

FALLACIES OF NARRATION

141

New England town library, there is likely to be an ancient Puritan
virgin, shriveled and dried in the snows of sixty Massachuetts win-
ters and suitably shrouded in black bombazine, who has been at work
for the past twenty years on the story of her home town from 1633 to
1933, when Franklin Roosevelt was inaugurated and history came to
an end. In the Maryland Historical Society one might find a retired
colonel, impeccably dressed, with the Bronze Star in one buttonhole
and the rosette of the Company of Military Collectors in the other,
who is writing a monograph on the smallclothes of the Continental
Army. At the American Jewish Historical Society there may be an
elderly gentleman at work on an article called "A Jewish Tourist at the
Battle of Bladensburg." In the New York Public Library there may be
a desiccated country clergyman, in black oxfords and white athletic
socks, who is at work on a county history of his denomination, from King
William's War to the Peace of Ghent, in two octavo volumes, illustrated.
Three years ago, the University of Oklahoma Press published a book
by a petroleum geologist who is said to possess the world's largest private
collection of barbed wire, which he apparently treasures in the proper
antiquarian spirit. 21

But the antiquarian fallacy is not confined to antiquarians. When-
ever a professional historian deliberately tries to cut himself off from
his own time in order to study somebody else's, he commits the same
mistake. Winthrop Jordan, author of a splendid scholarly work on white
American attitudes toward Negroes before 1812, is an example. "I have
attempted, indeed, to avoid reading widely in the literature of the present
crisis because it is frequently so tempting to read the past backwards —
and very dangerous," he writes. 22 One may sympathize with Jordan's
desire to avoid the presentist fa'lacy, but deliberate ignorance is no
defense.

Many other historians (the number is diminishing) have argued
that a history of ongoing events ought not to be attempted, because
objectivity is impossible, evidence is incomplete, and perspective is
difficult to attain. 23 But all these complaints also apply to ancient history,
if not precisely in the same degree. Most reasonable men would agree,
I think, that the history of contemporary events presents some special

21. Francis T. and Henry D. McCallum, The Wire That Fenced the West (Norman,
Okla., 1965).

22. Winthrop D. Jordan, White over Black (Chapel Hill, N.C., 1968), p. ix.

23. Geoffrey Barraclough discusses this controversy, which has been going on for forty
years and more in the English journal History, in An Introduction to Contemporary
History (New York, 1965), pp. 11-20.

142

EXPLANATION

difficulties but many other opportunities. To rule out events because
they are living is as mistaken as the opposite error.

The fallacy of tunnel history is identified by J. H. Hexter,
and so-named after the tendency of many historians to

split the past into a series of tunnels, each continuous from the remote
past to the present, but practically self-contained at every point and sealed
off from contact with or contamination by anything that was going on in
any of the other tunnels. At their entrances these tunnels bore signs saying
diplomatic history, political history, institutional history, ecclesiastical
history, intellectual history, military history, economic history, legal history,
administrative history, art history, colonial history, social history, agricultural
history, and so on, and so on. At first glance one might think that these
kinds of history came into being as a consequence of a rational attempt at
an exhaustive classification of what is knowable about the past, and that
history continues to be written under these headings because the classifi-
cation represents the best way to deal with the past. Nothing could be
further from the truth. Whether or not the current classification provides
the best means for exploring the past, it did not come into being as a
result of any conscious aiming at the best on the part of historians.

What mainly determined the way historians split up history during the
past century was a ridiculously adventitious set of circumstances: the way
in which public authorities and private persons tended to order the docu-
ments which suited their purposes to preserve. The basis of constitutional
history was the preservation and segregation of the commands of the
superior public authorities, of administrative history the care with which
accountable public servants preserved the documents for which they were
responsible, of intellectual history the irrepressible urge of intellectuals
and literary folk to hand on their maunderings to posterity and to display
their erudition by references to the writings of other intellectual and literary
folk, and so on. 24

In terms of a hypothetical historical matrix, tunnel history takes
a problem which happened thus:

Time Past events

1 A B C D E

2 B C D E F

3 C D E F G

4 D E F G H

5 E F G H I

and converts it into

(Past)
(Present)

24. J. H. Hexter, Reappraisals in History (Evanston, 111., 1961), pp. 194-95.

FALLACIES OF NARRATION

143

Time Past events

(Past) 1 A

2 B

3 C

4 D
(Present) 5 E

or maybe

Time Past events

(Past) 1 E

2 E

3 E

4 E
(Present) 5 E

Each of these versions of the past is not merely incomplete but seriously
inaccurate, which is what happens when a complex problem of develop-
ment is taken apart and its components are extruded into long thin
ribbons of change.

Unpleasant things are apt to happen when one grimy historical
tunneler bumps into another, somewhere in the dark, and disputes the
right of way with pick and shovel. In these altercations, the fallacy of
essences is apt to be the ultimate weapon. Charles Beard, for example,
insisted that the United States Constitution was "essentially an economic
document." To this, Henry Steele Commager made the absurd reply
that it "was, and is, essentially a political document." 25 Surely the Con-
stitution was in some respects an economic document and in other
respects a political document, and essentially neither, and indeed
essentially nothing.

The same delusion is much more dangerous when it appears in
arguments over American strategy in Vietnam. Some insist that a
successful policy must be "essentially political" or "essentially social" or
"essentially military" or "essentially educational" or "essentially eco-
nomic" or "essentially cultural." Again, the tunnelers (few of whom, in
this instance, are academic historians) meet in the dark and fight it
out, and everybody loses. 2 *

25. Charles A. Beard, An Economic Interpretation of the Constitution (New York,
1935), p. 324; Henry Steele Commager, The Search for a Usable Past (New York.
1967), p. 60.

26. Cf. Hexter, Reappraisals in History, p. 199.

144

EXPLANATION

There are many different kinds of tunnels in historiography. Among
the narrowest and darkest are the ethnic tunnels. And of all the ethnic
tunnels, none is quite so dark and narrow as that which is called "Jewish
history." There are many important and even urgent things to be written
about the history of the Jewish people, but the present mode of writing
and research is a scandal and an abomination in its profound provincial-
ity. Jewish historiography converts a matrix such as:

Time Past events

(Past) 1 F G H I J

2 G H I J K

3 H I J K L

4 I J K L M
(Present) 5 J K L M N

into:

Time Past events

(Past) 1 J

2 J

3 J

4 J
(Present) 5 J

The American Jewish Historical Society, for instance, dutifully sub-
scribed for many years to a large number of historical journals, some
of which are rare today. But the librarians of that institution believed
it their duty to snip out everything in the journals which explicitly
referred to Jews and to throw away the rest. One might doubt that any
chosen people are quite as chosen as all that.

It ought to be apparent that people who are commonly called
Jews by reason of their membership in an ethnic-religious group are
also at the same time members of many other groups. No problem in
their history can be resolved within the single aspect of their Jewishness.
This fact, which many Jews and non-Jews alike are reluctant to acknowl-
edge, is the great antidote to the poison of anti-Semitism which has
caused so much suffering and hatred and death in the world.

$»» The fallacy of false periodization consists in assigning inap-
propriate temporal limits to a historical problem. It is sometimes asserted

FALLACIES OF NARRATION

145

that all periodization is necessarily false and artificial. This, I think,
is incorrect. Maitland's famous epigram that history is a "seamless web"
betrays the holist bias of that great historian. History is a web of many
seams. The problem is not that a scholar must make one out of whole
cloth, so to speak, but that he must choose one out of many, and that
he may make a mistaken choice.

One common kind of false periodization might be called hecto-
history. It happens when history is neatly chopped into Procrustean
periods, each precisely a hundred years long. The fascination of a
rounded number is irresistible. "We tend to count by centuries," Marc
Bloch complained.

We no longer name ages after their heroes. We very prudently number
them in sequence every hundred years, starting from a point fixed, once
and for all, at the year 1 of the Christian era. The art of the thirteenth
century, the philosophy of the eighteenth, the "stupid nineteenth": these
faces in arithmetical masks haunt the pages of our books. Which of us
will boast of having never fallen prey to the lures of their apparent
convenience? 27

This species of distortion may be particularly serious in France,
owing to an accident of language, by which the single word siecle has an
unfortunate double meaning — both "century" and "period." English-
speaking historians are a little luckier in this respect. Nevertheless, the
fallacy of hectohistory often appears in the United Kingdom and the
United States. What Thomas Hardy called the "centuried years" are
commonly yoked together: "Each century seen in perspective, has its
own unique pattern. The thirteenth century was one of deep contrasts:
of consummate skill and real elegance side by side with barbaric and
primitive ignorance; of cruelty and corruption on the one hand and
saintly courage on the other." 28

Hectohistory is undoubtedly useful to undergraduates, as it was
to T. E. Lawrence, sometime of Arabia. It is reported that "In his
History Finals at Oxford he spoke merely in centuries — 'in the beginning
of the twelfth' or 'towards the end of the fifteenth century,' etc. Only
once he named a date — in brackets: 'about the middle of the eleventh
century (1066).' . . ," 29

The same generic form of error has imposed itself on history in

27. Marc Bloch, The Historian's Craft (New York, 1953), pp. 181-82.

28. T. Beamish, Battle Royal: A New Account of Simon de Mont fort's Struggle Against
King Henry III (New York, 1966), p. 31; for another example see Henry Horowitz,
Revolution Politicks: The Career of Daniel Finch, Second Earl of Nottingham (Cam-
bridge, 1968), p. vii.

29. Sir Lewis Namier, "Lawrence: As I Knew Him," In the Margin of History
(London, 1939), p. 279.

146

EXPLANATION

many other forms. More than a few Christian scholars have made most
things divisible by three. Others have been obsessed with a periodization
based upon the four monarchies in the prophecy of Daniel (2:40). Still
others have divided the past into six ages, according to the six days
of creation. A few medieval scholars sundered things seven ways, in
proportion to the seven planets. In Vietnam, nine is a lucky number.
Among astrologers, twelves seem more natural than tens. 30

A different form of false periodization appears when a historian
takes a time scheme which may be valid and functional in problem A
and transfers it to problem B, where it is invalid and dysfunctional.
American history is still periodized in the textbooks by presidential ad-
ministrations, which are perfectly proper for a history of the presidency
but not for the development of American society, which possesses its own
set of inaugurations and retirements. 31 This unfortunate method of
periodization is quite as primitive as that which traditionally prevailed
in the oral history of the Congo, where history was reckoned in rainy
seasons and dry. 32

Another kind of periodization error is one in which the historio-
graphical medium provides the method. A classic example is that species
of quasi-historical writing which appears in newspapers and newsmag-
azines. Academic historians are often contemptuous of the historical
interpretations of journalists — and properly so. Those failings are attribut-
able not to the cultural barbarism of the fourth estate, but rather to the
scheme of periodization which is forced upon it. Time, for a newspaper-
man, is measured in the intervals between editions. His often desperate
effort to find some significant happening in each of these periods explains
his shallowness, rather than ignorance or illiteracy or the company he
keeps.

A fourth kind of periodization error is a kind of temporal reduction,
in which a very large temporal period scheme is hung upon a very small
peg. A famous example is Michelet's division of the reign of Louis XIV
into two periods, "avant la fistule" and "apres la fistule" — a great epoch
in the history of a great nation, periodized by a painful anal fistula on
the bottom of the man at the top. 33

30. Wallace K. Ferguson, The Renaissance in Historical Thought (Cambridge, 1948),
p. 6, passim; Philippe Aries, Centuries of Childhood (New York, 1962), pp. 18-25.

31. Thomas C. Cochran, "The 'Presidential Synthesis' in American History," American
Historical Review, 53 (1948): 748-59.

32. Jan Vansina, Oral Tradition (Chicago, 1965), p. 100.

33. G. J. Renier, History, Its Purpose and Method (New York, 1965), p. 118, notes that
this interpretation "can be disproved through the study of the diary kept by the physicians
of the Grand Monarque. This manuscript, which was published in the nineteenth century,
reveals that the health of Louis had been extremely bad for many years before a fistula

FALLACIES OF NARRATION

147

The telescopic fallacy makes a long story short. It appears
in interpretations which reduce an extended trend to a momentary trans-
formation. This form of error is common today, and likely to become
still more so, as historians become increasingly interested in putting
big questions to little tests. When they do so, they must be careful not
to attempt to find a solution to the entire problem within the narrow range
of the period tested.

Three generalized examples should suffice. First, there has been a
good deal of loose talk about the historical revolution, i.e., the moment
of a great expansion in historical consciousness. The historical revolution
has been located by various monographers in ancient Sumer, ancient
China, ancient Israel, ancient Greece, and ancient Rome; in the thought
of a miscellany of medieval Christians, Jews, and Arabs; in the twelfth-
century Renaissance; in fifteenth-century Italy, seventeenth-century Eng-
land, eighteenth-century France, nineteenth-century Germany, and various
twentieth-century movements. There was probably a development of his-
torical consciousness in all these periods and places. But the development
of historical consciousness occurred in none of them. Any attempt to
locate a protracted trend in a brief period, as, for instance, F. Smith
Fussner does in The Historical Revolution: English Historical Writing and
Thought, 1580-1640 (New York, 1962), is fallacious.

Another example is the problem of the decline of deference in English
society. Lawrence Stone located this extended development in the early
seventeenth century. Virginia Woolf , on the other hand, placed it in the
twentieth century with remarkable precision.

In or about December, 1910 [she wrote], human character changed, I am not
saying that one went out, as one might into a garden, and there saw that
a rose had flowered, or that a hen had laid an egg. The change was not
sudden and definite like that. But a change there was nevertheless; and
since one must be arbitrary, let us date it about the year 1910. The first
signs of it are recorded in the books of Samuel Butler, in The Way of All
Flesh in particular; the plays of Bernard Shaw continue to record it. In
life one can see the change, if I may use a homely illustration, in the
character of one's cook. 34

Other scholars have dated the decline of deference in still other
ways — in eighteenth- or nineteenth-century England, for example. It
seems reasonable to guess that so extraordinary a transformation oc-
curred gradually (not steadily) over many generations. Maybe human

analis was diagnosed, and that this new illness was not an exceptional phenomenon in the
career of this coarse contemner of hygiene and sensible living."

34. Virginia Woolf, "Mr. Bennett and Mrs. Brown," Collected Essays, 4 vols. (London,
1966), 1:321.

148

EXPLANATION

character did change a little, in December, 1910, but also in 1810, and
1710. One wonders if trend can be stretched back as far as 1610 without
breaking. The apparent intensity of deference politics in late seven-
teenth-, eighteenth-, and nineteenth-century England would appear to
contradict Stone's thesis. But let us give him the benefit of the doubt
and assume that there was, indeed, something like a decline of deference
in his period. The next question concerns the relationship of that decline
to developments in other periods. Stone does not help us find the answer.
He telescopes the problem into his period, and, except for a few brief
and inconclusive passages, he ignores other eras. 35

A third academic example of the telescopic fallacy occurs in a book
called The Strange Career of Jim Crow, by C. Vann Woodward. 36 The
author argues for the "relative recency" of the establishment of racial
segregation in the South — specifically that "bonds of intimacy" existed be-
tween the races during slavery, that there was a period of "forgotten
alternatives" after the Civil War, and that the "capitulation to racism"
occurred in the South during the 1890s and afterward. Woodward's
purpose, apparently, was to demonstrate to Southerners that their al-
legedly "immutable folkways" were in fact modern mutations, which,
having changed once recently, could change again.

This thesis was greeted with enthusiasm by many of Woodward's
colleagues — some of whom proclaimed to the world that "racial segrega-
tion in the Old South had been unknown," and that Jim Crow "did not
spring directly from slavery or from the timeworn customs of many
generations," and that before the Civil War, "the circumstances which
later gave rise to the segregation codes could not exist." 37

These unguarded statements, and the more moderate thesis from
which they stem, are classic examples of the telescopic fallacy. Much
recent work has demonstrated in great detail that Woodward's argument
is wrong in all its major parts — wrong because he made a long story
short. There is abundant evidence of a "capitulation to racism" by
Americans, both above and below the Mason-Dixon line, long before
1890. No period of "forgotten alternatives" existed after the Civil War —
racial segregation was immediately imposed by law and custom. Even
before the war, segregation formally existed wherever free Negroes could

35. Lawrence Stone, The Crisis of the Aristocracy: 1558-1641 (New York, 1965),
pp. 7-17, 746-53 passim.

36. (Oxford, 1955).

37. Barton J. Bernstein, "Plessy v. Ferguson: Conservative Sociological Jurisprudence,"
Journal of Negro History 48 (1963): 200; John S. Ezell, The South Since 1865 (New
York, 1963), p. 184; Charles Crowe, ed., The Age of Civil War and Reconstruction
(Homewood, 111., 1966), p. 439; quoted in Roger A. Fischer, "Racial Segregation in
Ante Bellum New Orleans," The American Historical Review 74 (1969): 926-27.

FALLACIES OF NARRATION

149

be found. And most important, slavery functioned in the Southern states
not merely as a labor system but also as an instrument of social control
which was itself a form of racial segregation (physical as well as social)
more extreme in many ways than the most elaborate Jim Crow codes. In
short, the history of racism and racial segregation in Anglo-America is
a very long story indeed. Woodward telescoped it into the narrow span
of his own special period, and thereby falsified it. 38 There were un-
doubtedly important changes in the quality of racist thought, and in the
shape of racist institutions, during Woodward's narrow period — changes
which affected a far larger area than merely the South. But the fact
remains that the thesis of the Strange Career of Jim Crow truncates one
of the enduring themes of American history. 39

This form of the telescopic fallacy — in which distant origins are
forgotten — is its most common form. It is a serious problem in historians'
sources as well as in their own writings. Survey research has yielded solid
evidence that respondents often tend to telescope their retrospections.
Historians face similar problems in retrospective accounts of events. 40

Precisely the same error often occurs today in discussions of con-
temporary problems. How many commentators on the consequences
of "automation" have assumed that the introduction of labor-saving
machinery is something which our generation is the first to face? It is
an incredible error, and yet one which is very common and very delete-
rious in its influence upon strategies designed to deal with the problem.
Sometimes it makes a mighty difference to know how long ago something
began to happen.

$»» Conversely, the interminable fallacy makes a short story long,
or a long story longer than it ought to be. It is a temporal form of a

38. Among many recent revisionary works, see Joel Williamson, After Slavery (Chapel
Hill, 1965); Richard C. Wade, Slavery in the Cities: The South, 1820-1860 (New York,
1964); Leon Litwack, North of Slavery (Chicago, 1961); V. Jacque Voegeli, Free but
Not Equal: The Midwest and the Negro During the Civil War (Chicago, 1967); and
Winthrop Jordan, White over Black (Chapel Hill, 1968).

39. Through two revisions, the author has held his ground with a tenacity worthy of
a better cause. The result is another fallacy — the overwhelming exception (see above,
chapter 4). We are now told that the interpretation applies to all Southern institutions
except churches, schools, militia, hotels, restaurants, public buildings, jails, hospitals,
asylums, gardens, railroads in several states, and the New Orleans Opera House. It
applies to all Negroes but freedmen, hired slaves, urban Negroes, South Carolina
Negroes, and in some respects Negro field hands (1957 ed., p. 15; 1966 ed., pp. 12,
13-17, 25-26).

40. John Neter and Joseph Waksberg, "A Study of Response Errors in Expenditures
Data from Household Interviews," Journal of the American Statistical Association 59
(1964): 18-55.

150

EXPLANATION

false extrapolation — a developmental trend stretched beyond the breaking
point.

A familiar example is that omnipresent cliche of modern European
historiography, the "rise of the middle class." This group has been found
rising most remarkably in every period from the twelfth century to the
twentieth. It has been used to explain the Renaissance and the Reforma-
tion, absolutism and liberalism, monarchy and republicanism, conserva-
tism and radicalism, nationalism and internationalism, romanticism and
rationalism, fascism and communism, the commercial revolution, the
managerial revolution, the agricultural revolution, the industrial revolu-
tion, the Puritan revolution, the American Revolution, the French Revo-
lution, the Russian Revolution, etc., etc.

If the middle class had in fact been rising as powerfully as this, it
should presently be somewhere in the disciplinary jurisdiction of astron-
omers, who alone could measure its continuous ascension with their
powerful instruments.

This risible phenomenon surely requires no extended refutation.
But the interested reader might examine an unfortunate book by Louis
B. Wright called Middle-Class Culture in Elizabethan England (Chapel
Hill, 1935), and an excellent critique by J. H. Hexter, called "The Myth
of the Middle Class in Tudor England." Wright found a close interactive
causal connection between the rise of the middle classes and the rise of
the Tudors, an allegedly "bourgeois" dynasty. Hexter correctly complains
of the tautological nature of this thesis, in which everybody who appears
to be rising is admitted to the middle class. Hexter wonders not merely
whether or not the middle class was rising in Tudor England, but
whether it was there at all, in any meaningful sense. He finds that
"there is little evidence, then, that the Tudor period saw any extraordinary
development in the middle class of group conciousness, group pride, or
will to power," and no proof of the proposition that Tudor monarchs
favored commerce in any special way, or that they manifested "middle-
class" characteristics in any intelligible sense. 41

The fallacy of archetypes consists in conceptualizing change
in terms of the re-enactment of primordial archetypes which exist out-
side of time. It is a method in which an event acquires meaning as a
re-enactment of some aboriginal and atemporal model. Mircea Eliade
supplies many examples, all of which manifest the same primitive onto-
logical conception — that "an object or an act becomes real only insofar

41. Hexter, Reappraisals in History, p. 99.

FALLACIES OF NARRATION

151

as it imitates or repeats an archetype. Thus, reality is acquired solely
through repetition or participation; everything which lacks an exemplary
model is 'meaningless,' i.e., it lacks reality. Men would thus have a ten-
dency to become archetypical and paradigmatic. This tendency may well
appear paradoxical, in the sense that the man of a .traditional culture
sees himself as real only to the extent that he ceases to be' himself." 42

In archaic societies, the myth of the eternal return to aboriginal
archetypes may be merely "paradoxical," as Eliade calls it. But when the
myth becomes a modern interpretation of historical change, the paradox
becomes a fatal contradiction. The myth of the return is an antithesis to
time, change, and history itself. When it is used by a historian to con-
ceptualize his subject, then it becomes a fallacy, for the myth implies
that what is real does not change. His time series is bent back upon itself
in a sterile series of cyclical enfoldments.

The classical examples of this fallacy are those monumental works
of melancholy genius, Arnold Toynbee's A Study of History and Oswald
Spengler's The Decline of the West. Both authors conceived all "civiliza-
tions" according to fixed archetypal patterns which transcend time.
There are many complaints to be entered against these interpretations.
But much of their absurdity flows from a sense of time which is a perfect
example of Mircea Eliade's idea of primitive prehistorical thought.

Toynbee's "civilizations" are all converted into conceptual con-
temporaries. When this is done to Sumerian "civilization" on the one
hand and Western "civilization" on the other, the result may be the most
gigantic anachronism in recorded historiography. The vast developmental
changes from one to the other are either ignored or discounted. Similarly,
between "Sumerian" and "Sinic" civilization, between "Minoan" and
"Christian" civilization, there are contrasts of complexity and scale which
are so great that any simple-minded archetypical interpretation is at
bottom absurd. 43

42. Mircea Eliade, Cosmos and History: The Myth of the Eternal Return (New York,
1959) pp. 34-43, passim.

43. There are also many small single examples of archetypical thinking in historiog-
raphy — examples which are significant only in a symptomatic sense. A petty specimen
appears in a fine history of Scotland by John Duncan Mackie, in which that excellent
author asserts that "history had repeated itself exactly . . . the year 1286 was come
again [in 1542]." Similarly, G. A. Williamson writes in another context, "No one can
read Eusebius's account of how the cathedral of Tyre, with all its elaborate symbolism,
rose from the ashes, without thinking of Coventry. Truly that generation and this are
one." Whenever these phrases appear before a reader's eye, a klaxon horn should
sound a warning in his ear. See J. D. Mackie, A History of Scotland, 2d ed. (Balti-
more, 1966), p. 141; and Eusebius, The History of the Church from Christ to Con-
stantine, ed. G. A. Williamson (New York, 1966), p. 10. Another example suggests
that the warning sound might well be a few bars from Le Retour d la Vie. See Jacques
Barzun, Berlioz and His Century (New York, 1956), p. 10.

152

EXPLANATION

The fallacy of archetypes is an erroneous form of historical con-
sciousness, but it is not restricted to the conciousness of erroneous
historians. There are many people in modern industrial societies who
still live in the grip of Mircea Eliade's "archaic ideology of ritual repeti-
tion" and dream of the return to the eternal archetype. In Cornwall,
today, there is a band of fanatics who seriously predict the second coming
of King Arthur. 44 Others dream of — who knows what? The second com-
ing of Charlemagne, or Christ, or Thomas Jefferson; the recovery of the
primitive church, or the garden of Eden, or some other Shangri-la for
which real people are ready to butcher other real people who seem to
stand in the way. Here is another fatal fallacy that cannot be allowed to
endure in a complex world.

The chronic fallacy is a kind of misplaced temporal literalism
in which a historian forces his story into an overrigid chronological
sequence and tells everything in the precise order of its occurrence, with
results that are dysfunctional to his explanatory purpose. It is not easy to
define this form of error exactly, for a historian must begin by respecting
the chronology of events. But sometimes he can be overly obsequious to
the tyranny of time.

One might imagine a better method. A scholar should seek the pat-
tern of change in his problem and organize his narrative around it, rather
than according to a strict observance of the calender. This pattern always
exists, but never coincides exactly with the ticking of a clock. Imagine,
for example, the history of generational responses to a given theme. An
older generation may clearly express its attitudes at time 1, 3, 5, 7, and 9.
A younger generation might do so at time 2, 4, 6, 8, and 10. It surely
makes better sense to tell this story in the sequence 1 , 3, 5, 7, 9, 2, 4, 6,
8, 10, and not in a straight numerical order, which would make literal
nonsense of the story.

If the chronic fallacy is not easy to define, it is unmistakable in its
commission. There are many examples in the work of great nineteenth-
century historians, who were trained to think narrowly in terms of linear
narration and never to take a liberty with a date.

But the error survives in our own era. Witness John A. Garraty's
biography of Henry Cabot Lodge, which tells everything with an engag-
ing simplicity but explains nothing in a satisfactory way. The complex
rhythms of Lodge's career are reduced to the mindless cadence of a
metronome. The resultant narrative structure is not a vehicle for ex-

44. George Thayer, The British Political Fringe (London, 1965), Chap. 9.

FALLACIES OF NARRATION

153

planation but a means of compiling 400 pages without one. 45 No satis-
factory sense of the quality of Lodge's conservatism is communicated to
the reader, nor any extended awareness of the psychic, social, and cul-
tural contexts within which Lodge lived. There is, I think, an important
developmental sequence in Lodge's public career — a measured develop-
ment through time, though not precisely Eastern standard time. There
is a kind of counterpoint of environment, attitude, and act, which imposes
its own chronology upon the problem. But in Garraty's book, Anno
Domini, dominus est.

The static fallacy broadly consists in any attempt to concept-
ualize a dynamic problem in static terms. This form of error represents
an intermediate stage of historical consciousness, in which change is per-
ceived merely as the emergence of a nonchanging entity. Such was the
shape and texture of much Christian historiography, which often repre-
sented past events in terms of a slow unfolding of a preordained divine
plan. The same conception, semisecularized, appears in George Ban-
croft's great history of the unfolding of the American plan. Bancroft's
chapter on the earliest Anglo-American settlements is titled, "England
Takes Possession of the United States." 46

Even in our own time, a good many liberal textbook historians of
the American republic tend to conceptualize their dynamic subject in
terms of the unfolding of a static idea of democratic society, which slowly
reveals itself through three centuries, without ever really changing in the
process. The result is a historiographical equivalent of the Dance of the
Seven Veils, featuring the damsel Democracy herself, and half a dozen
willing helpers. First, Roger Williams helps her out of a somber shroud
of Puritan black. Then Benjamin Franklin rends a red coat with his light-
ning rod, and Thomas Jefferson tugs off a covering of Hamiltonian buff
and blue, to expose an earthy homespun of Old Hickory brown. This
rude garment falls to pieces, revealing a cloak of Confederate gray,
which Lincoln removes with magnanimous gestures. Next there is a
gilded robe, embroidered with Black Fridays and costly touches of Tweed,
which miraculously yields to a checkered cloth of Populist red and Pro-
gressive lily white, with a free-silver lining. The last veil finally falls away,
and beauteous Columbia stands revealed, with a blue eagle tattooed on
her belly.

In all seriousness, any scholar who tries to organize a book called

45. John A. Garraty, Henry Cabot Lodge: A Biography (New York, 1953).

46. George Bancroft, History of the Colonization of the United States, 15th ed.
(Boston, 1852), 1:74.

154

EXPLANATION

The Growth of the American Democracy around the emergence of a
single static democratic idea will soon find himself tied up in conceptual
knots. If, on the other hand, the theme is shifted from the emergence of
democracy to democratization, and if the latter is understood not as the
maximization of an immutable idea, but rather as a series of fluid ideas
each of which are qualitatively in motion on several separate axes, then
at least a few of the obstacles in the path of historical understanding are
removed.

The number of "-izations" in the lexicon of history is rapidly in-
creasing, as our understanding of change is steadily refined. But the
static fallacy still appears in sophisticated disguises. American political
historians, for example, stand ready to argue at the drop of a cocked
hat, over the question "When did the first modern party system appear in
the United States?" Many find the answer in the Jacksonian era. A few
date the development during the 1840s. Some go back to the 1790s.
Others locate the first really modern party system in the Jeffersonian era.

This academic dispute, to which I regret that I have contributed,
is wildly unhistorical when it is cast in these terms. It derives from a
faulty question, which rests in turn on the false idea of a modern party
as something that exists statically outside the flow of time but slowly
unfolds within it. Any attempt to locate a single model of the modern
party system for purposes of historical analysis is not merely arbitrary,
but absurd.

$»» The fallacy of presumptive continuity and the fallacy of
presumptive change are two fundamental forms of error which came to
mind in a reading of Barrington Moore's Social Origins of Dictatorship
and Democracy. Moore criticizes his colleagues for an assumption of
social inertia in their work.

There is a widespread assumption in modern social science [he writes]
that social continuity requires no explanation. Supposedly it is not problem-
atical. Change is what requires explanation. . . . The assumption of inertia,
that cultural and social continuity do not require explanation, obliterates the
fact that both have to be recreated anew in each generation, often with
great pain and suffering. To maintain and transmit a value system, human
beings are punched, bullied, sent to jail, thrown into concentration camps,
cajoled, bribed, made into heroes, encouraged to read newspapers, stood
up against a wall and shot, and sometimes even taught sociology. To
speak of cultural inertia is to overlook the concrete interests and privileges
that are served by indoctrination, education, and the entire complicated
process of transmitting culture from one generation to the next. 47

47. Pp. 485-86.

FALLACIES OF NARRATION

155

As an example of an "assumption of inertia," Moore cites the work
of Talcott Parsons, 48 who does indeed explicitly commit this error. But
I wonder if Moore has committed the counterfallacy of presumptive
change. They are both equally indefensible. Probably no form of bias
is more difficult to eradicate from one's work than this one, and yet its
consequences can be exceedingly serious. One thinks of those two coun-
tervailing schools of American historiography — the consensus-continuity
school of B. F. Wright, Louis Hartz, and Daniel Boorstin, and the
conflict-change school of an earlier generation of progressive historians.
What is clearly needed is a set of mediating terms and concepts which
might help to neutralize these opposite biases.

$»» The genetic fallacy mistakes the becoming of a thing for the
thing which it has become. In other words, it is the erroneous idea that
"an actual history of any science, art, or social institution can take
the place of a [nontemporal] logical analysis of its structure." 4 "

Much was made of the genetic fallacy by the so-called New Critics
of the interbellum era — T. S. Eliot and others who invigorated their
discipline by insisting that a history of a literary work is something
different from an analysis of its aesthetic structure. A few went further
and argued that literary history is no part of a literary critic's business,
and no use to him in his work. This counterfallacy, which confuses a
thing-which-has-become for a thing's becoming, has been amply con-
demned by the so-called New-New Critics, such as Leslie Fiedler and
others. 50 Both kinds of organizing schemas — temporal and nontemporal
alike — are surely useful and constructive tools of analysis and explana-
tion. But they are not interchangeable. The New Critics were correct in
drawing a clear distinction between them.

The most hateful forms of the genetic fallacy are those which con-
vert a temporal sequence into an ethical system — history into morality.
This pernicious error was embedded in a movement called historicism,

48. The Social System (Glencoe, 1951), p. 205.

49. Morris R. Cohen and Ernest Nagel, An Introduction to Logic and the Scientific
Method (New York, 1934), pp. 388-90. I think it incorrect to distinguish, as Cohen
and Nagel do, between a temporal and logical order. A narrative is itself a special kind
of logical order. If so, the proper distinction is therefore a temporal logic on the one
hand and a nontemporal logic on the other.

To argue that narrative possesses a logical structure of its own is not, however, to
suggest that past happenings themselves possess an inherent logic of their own, but
rather that a study of past happenings must be conducted in a logical manner, and that
the resultant interpretation must be organized in a logical form if it is to be intelligible.

50. For an extended critique of the "new criticism" in this respect, see E. D. Hirsch,
Jr., Validity in Interpretation (New Haven, 1967).

156

EXPLANATION

which flourished in Germany during the period 1790-1930 — a school
which stretched from the early work of Herder, Hegel, Schiller, and
Schelling to the later work of Troeltsch and Meinecke. Historicism was
many things to many people, but in a general way its epistemology was
idealist, its politics were antidemocratic, its aesthetics were romantic,
and its ethics were organized around the nasty idea that whatever is
becoming, is right. 51

The classical expression of ethical historicism is Schiller's epigram
"die Weltgeschichte ist das Weltgericht." This doctrine reduced ethics
(and much else) to a province of historiography. And it was radically
destructive, not merely of other ethical systems but of itself as well.
Ethical historicism commonly took one of two untenable forms. Some
historicists — Ranke is an eminent example — unwittingly smuggled an
ethical system into history, and then discovered it as the objective teach-
ing of history itself. Others later converted ethical historicism into an
ethical relativism. Meinecke, for instance, asserted that "nothing can be
immoral which comes from the innermost individual character of a
being." 52 This doctrine must necessarily become an ethical nihilism. It
would prevent any moral judgment against the filth which flowed from
"the innermost individual character" of many Nazi beings. Historicism,
relativism, nihilism. There is no stopping place in this downward descent
to nothingness. 53

German historicism is dead, or dying, but the same ethical version
of the genetic fallacy still appears in other forms. It seems to find a
certain popularity in what the authors of 7066 and All That called
"Top Nations." American historians such as Daniel Boorstin came close
to arguing in the 1950s that Die americanische Geschichte ist das
Weltgericht, and they were not alone in that assumption. Something of
the fallacy of ethical historicism appears in the absurd and dangerous
idea that America's rise to power and prosperity is a measure of its
moral excellence — that the history of the Republic can be seen, in short,
as a system of morality. How many of us have not, at some time, silently
slipped into this error?

51. The many conflicting (and sometimes contradictory) definitions of historicism
are a measure of the movement's range and complexity. The best and fullest account
in English is Georg G. Iggers, The German Conception of History (Middletown,
Conn., 1968), with a good bibliography.

52. Weltburgertum una" Nationalstaat, in Werke (Munich, 1959), 5:83; as quoted in
Iggers, The German Conception of History, p, 271.

53. The doctrines of historicism are sometimes condemned as the cause of Nazism,
as if Herder puffed up a great cloud of Teutonic humbug and out marched Adolf
Hitler. This is unfair, inaccurate, and illogical: a misreading of German history, in
my opinion; and a classic example of an argument to consequences. Historicism
remains profoundly hateful, but for itself, and not for its putative effects.

FALLACIES OF NARRATION

157

On the other side of the ideological divide, another contemporary
variant of the fallacy of ethical historicism might be diagnosed as Carr's
disease, after an English socialist scholar who seems to think that
morality marches triumphant through history, always on the side of the
big battalions. Carr marches through history with them, too. He is not
a citizen of a Top Nation, but a strenuous advocate of a Top Notion.

One wonders how Carr measures the bigness of his big battalions —
a tough perception problem which cannot be resolved merely by the quan-
tification of muster rolls. Psychologists have demonstrated that a symbol
of something which is valued looks physically bigger than a symbol of
something which is not. 54 Their discovery applies to Carr's big battalions,
and to much else besides. In that degree his argument becomes circular,
as all ethical historicisms are. An ethical system is bootlegged into
history, and then proclaimed to the world as the lesson of history itself.
Or else, all ethical systems are undercut, so that nothing remains. The
only remedy is a radical recognition that a temporal series cannot be
converted into an ethical system, under any circumstances, without
self-contradiction .

The didactic fallacy is the attempt to extract specific "lessons"
from history, and to apply them literally as policies to present problems,
without regard for intervening changes. There are several special types
and subdivisions of this fallacy — e.g., the attempt to revive past precepts,
or creeds, or codes for the purification of the present; or the attempt
to repeat past political successes by returning to the programs of success-
ful politicians; or the attempt to study problem A and to apply con-
clusions obtained from it to problem B, though B may be very different
from A, and far removed in time and place from A.

Many pundits today are in the habit of misquoting Santayana's
epigram, "Those who cannot remember the past are condemned to repeat
it." 55 Maybe some people have come to grief this way, but they are
probably fewer than those who have fallen into the opposite error. "One
is apt to perish in politics from too much memory," Tocqueville wrote
somewhere, with equal truth and greater insight.

This is not to argue that historical study itself is useless or danger-
ous. It is not to endorse Hegel's maxim that "the only thing one learns

54. J. Bniner and L. Postman, "Symbolic Values as an Organizing Factor in Per-
ception," Journal of Social Psychology 27 (1948): 203 ff. Carr's position appears in
What Is History?, chaps. 2, 5, and 6; and The New Society (New York, 1957), passim.

55. George Santayana, The Life of Reason: Reason in Common Sense (New York,
1905), p. 284.

158

EXPLANATION

from history is that nobody ever learns anything from history." It is
rather to suggest that the utility of historical knowledge consists, among
other things, in the enlargement of substantive contexts within which
decisions are made, and in the refinement of a thought structure which is
indispensible to purposeful decision making about men, societies, time,
and change — a kind of logic which this book is an effort to define.
Historical study, in short, equips men to think historically. That intel-
lectual discipline is the only remedy against such errors as the following.

Consider the case of Field Marshal Douglas Haig, a Scottish soldier
who was centrally responsible for the bloody shambles at Passchendaele
in 1917. Some critics have attributed his costly blunders to sheer igno-
rance and stupidity. But that is not correct. Haig's mistakes were errors
of intelligence and learning. An able military historian has observed
that Haig

was not an uneducated soldier. Unlike so many cavalrymen of his day, he
had studied war, and, strange to say, this was to be his undoing, because
he was so unimaginative that he could not see that the tactics of the past
were as dead as mutton. We are told he held that the "role of cavalry
on the battlefield will always go on increasing," and that he believed
bullets had "little stopping power against the horse." 56

Other examples appeared in the next war, when, as always, many
soldiers and statesmen were superbly prepared to refight the last one.
There is a continuing controversy about Hitler's decision to halt his
panzers when they were but a few miles from Dunkirk. Chester Wilmot
provides a sensible explanation in his excellent history of World War II.

The Dunkirk decision [he wrote] was Hitler's first great military mistake,
but it was not so irrational and short-sighted as some German generals have
asserted. Although Hitler was influenced by the lure of Paris, the goal of
all German conquerors, a more important factor was his determination to
avoid the costly mistakes of 1914. He would not falter and be stopped
at the Somme, as von Moltke had been stopped on the Marne through
excessive concern about the British on the flank. 57

On the other side, General George S. Patton prepared himself in a
curious way, on the eve of overlord, for his extraordinary campaign
against the Wehrmacht in 1944.

I also read The Norman Conquest by Freeman [he wrote in his Diary]
paying particular attention to the roads William the Conqueror used in
his operations in Brittany and Normandy. The roads used in those days
had to be on ground which was always practicable. Therefore using these

56. J. F. C. Fuller, Introduction to Leon Wolff, In Flanders Fields (New York, 1960),
p. 9.

57. Chester Wilmot, The Struggle for Europe (New York, 1952), p. 21.

FALLACIES OF NARRATION

159

roads, even in modern times, permits easy by-passing when the enemy
resorts, as he always does, to demolition. 58

Fortunately for Allied arms, General Patton seems not to have had an
opportunity to apply these "lessons of history" very closely. Had he
done so, it is conceivable that the Third Army might still be hacking away
at the bocage.

More specimens of the same sort of error appear in political
history. Franklin Roosevelt during World War II appears to have been
dangerously obsessed with the failures of Woodrow Wilson in World
War I. 5a De Gaulle, who was at one time a professor of history at St.
Cyr, may have learned some "lessons" of French history all too well.
Stalin seems to have done the same with respect to the "lessons" of the
history of Russian foreign relations. And Winston Churchill, at Cairo in
1943, had a haunting conversation with Harold Macmillan. Late one
night, Churchill observed, "Cromwell was a great man wasn't he?"

"Yes, sir," said Macmillan, "a very great man."

"Ah," Churchill said, "but he made one terrible mistake. Obsessed
in his youth by fear of the power of Spain, he failed to observe the rise
of France. Will that be said of me?"

Churchill was, of course, referring to Germany and Russia. But
the opposite is said of him, increasingly, these days — that he was ob-
sessed too much with Russia, with serious consequences for the postwar
world. 60

The didactic fallacy is particularly dangerous to conservative
thought. One case in point is an unusually intelligent, and much maligned,
American president, Herbert Hoover. A historian, Herbert Feis, fairly
remarks,

President Hoover was not an insensitive or inhumane man; quite the
contrary. But he could not grasp or would not face the grim realities which
called for deviations from principles and practices that he deemed essential
to American greatness and freedom. The policies and proposals which he
expounded so earnestly might have served to end the depression, let us
say, in the 1870's or 1880's, but not the ones by which the United States
was then [after 1929] beset. 61

The same sort of mistake is contained in the atomistic conservatism
of Barry Goldwater's Conscience of a Conservative. Goldwater is an
intelligent and rational man, though the nature of his intellect and

58. George S. Patton, Jr., War as I Knew It (Boston, 1947), p. 92.

59. Eric Goldman, Rendezvous with Destiny (New York, 1952), pp. 398-400, 447-50.

60. Randolph S. Churchill, Winston S. Churchill, The Young Statesman (Boston, 1967),
p. 274.

61. Herbert Feis, 1933: Characters in Crisis (Boston, 1966), p. 7.

160

EXPLANATION

rationality is so far removed from that of many American academicians
that they are only able to understand him by presuming some degree
of insanity. He has made a very great rational error in his thought, in an
attempt to revive precepts and principles that were functional in nine-
teenth-century America but which are today dysfunctional and even
dangerous to the peace and prosperity of the world, for they ignore great
demographic, technological, social, and cultural changes. Walter Bagehot
has truly written that "the whole history of civilization is strewn with
creeds and institutions which were invaluable at first and deadly after-
wards."

£»> What positive principles of narrative explanation can be ex-
tracted from these fallacies? A few random thoughts come to mind.

First, a historian must distinguish between an analysis of the be-
coming of an object and an analysis of the object as it has become. The
narrative history of the writing of a poem is something distinct from an
analysis of its poetic structure. A narrative history of ethics is something
different from an ethical system. By the same token, a narrative history
of events in Southeast Asia is something other than a structural analysis
of the situation which presently exists. This lesson is not an easy one
for a historian to learn, for it requires him to recognize the limits of his
own expertise. And its converse is equally unwelcome to nontemporal
structural analysts of various persuasions. Some questions which we
ask about the world require explanatory answers of a narrative-descrip-
tive sort. Other questions require a different strategy. There is nothing
to be gained by a condemnation of either method out of hand — and much
to be lost by their confusion.

Second, the analysis of the becoming of a thing must be cast in
appropriately dynamic conceptualizations. Historians for many years at-
tempted to analyze dynamic problems in static terms, with seriously
dysfunctional results. Many scholars are slowly learning to employ more
fluid and flexible conceptual units. Conceptual flexibility is not conceptual
imprecision, but rather precision of a new kind, in which the dynamics
of change and continuity are carefully specified but not frozen into a
static structure. The stuff of history is things that happen — not things
that are. The historian's object, in Ranke's classic phrase, is not to tell
what actually was, but what actually happened.

Third, the logic of any narrative scheme must conform to the logic
of the problem at hand and not to some extraneous structure. Many dis-
tractions must be resisted. Calendars and clocks are the worst of tern-

FALLACIES OF NARRATION

161

poral tyrants, which reduce narrative histories to slavish chronicles.
The institutional structures within which historians live — academic de-
partments in America, for example — supply other snares which Hexter
has described in his attack on tunnel history. Antitemporal archetypes
still take their toll of serious research, and so also do many irrelevant
schemes of periodization.

Fourth, within an appropriate time scheme, events must be located
with accuracy and precision. A historian must preserve an uncompromis-
ing respect for the temporal integrity of his story and of its various
components. He must beware of the temptations of retrospective sym-
metry, in which antecedents are defined in terms of consequents. And
he must be on guard against anachronisms in every form.

Fifth, a historian must develop interpretative devices which neutralize
bias toward continuity or change. A device has been created by an
imaginative historian of science, Thomas S. Kuhn, from whom historians
in every field have much to learn. The history of science, in the past,
resembled an Old Testament genealogy. Copernicus begat Tycho, and
Tycho begat Kepler, and Kepler begat Newton, and the world was filled
with light. Kuhn has worked out a more satisfactory method, which he
has described in The Structure of Scientific Revolutions ( Chicago, 1962).

Kuhn's conceptualization of change and continuity in the history of
science is organized around a series of "shared paradigms" — sets of
scientific methods and understandings which were consistent with what
was known when they were established, and which persisted until anoma-
lies generated a crisis. The old paradigm was blurred, and then dis-
rupted; then a new paradigm was generated, and the process began
again. Kuhn writes:

Historians of science have begun to ask new sorts of questions and to
trace different and often less than cumulative, developmental lines for the
sciences. Rather than seeking the permanent contributions of an older
science to our present vantage, they attempt to display the historical
integrity of that science in its own time. They ask, for example, not about
the relation of Galileo's views to those of modern science, but rather about
the relationship between his views and those of his group, i.e., his teachers,
contemporaries and immediate successors in the sciences. Furthermore,
they insist upon studying the opinions of that group and other similar ones
from that viewpoint, usually very different from that of modern science —
that gives those opinions the maximum internal coherence and the closest
possible fit to nature. 82

62. Kuhn, The Structure of Scientific Revolutions, p. 3. One might compare Kuhn's
paradigms with a distinguished piece of retrospective symmetry, C. C. Gillespie's
The Edge of Objectivity (Princeton, I960), to see what a difference a method can
make.

162

EXPLANATION

Narrative history is still consistent with Kuhn's paradigms, but it
becomes a more profound and more intricate narrative, in which the
story consists not in a progressive unfolding of the present, but rather a
series of structural reformations (in a literal sense). Kuhn's method
may be marred by an excessive reaction to the linearity of earlier efforts,
but it is a closer approximation to the requirements of the subject than
anything we have had before. And it is relevant to all fields of historical
inquiry. Political history can be constructively conceived in precisely
the same paradigmatic terms — a certain congruence is established in a
polity, and then eroded by a sequence of new problems and purposes.
The old polity is at last overturned, there is a period of confusion, a
new polity is created, which possesses a congruence of its own, and the
process begins again The cyclical implications of this approach need
not be taken very seriously, for the nature of the paradigmatic relation-
ship is itself in motion, and it must vary greatly from one phenomenon
to another. To accept Kuhn's model is not to argue that history itself
is a right- or left-handed spiral, a sine curve, a merry-go-round, a roller
coaster, or a loop-the-loop. The method can coexist with any of these
absurd abstractions; it entails none of them. Instead, it is a flexible,
empirical device which means not the end of narrative history but the
beginning of a new kind of narration, which is in turn capable of fur-
ther refinement.

There are many problems of narration which have not even been
opened in this chapter — problems which nobody has managed to re-
solve. To tell a story is to assemble a group of mutually relevant facts.
But what is involved in the idea of relevance? "There are few problems
in philosophy which merit closer analysis than the question of relevance,"
Arthur Danto has written. 83 And few have received so little attention.
Now that relevance has become the rallying cry of a younger generation,
perhaps it will be properly investigated. 94

Narrative also involves the idea of connectedness among relevant
events. But what is the nature of connections in a narrative series?
"Tended, grew out of, developed, evolved, trend, development, tendency,
evolution, growth," J. H. Hexter has written. "Such words are like
sealed junction boxes on the complex circuits of history. One knows
that inside the boxes there are connections which induce the currents of
history to change direction; but the boxes conceal rather than reveal
how these connections are made." 90

63. Danto, Analytical Philosophy of History, p. 132.

64. For some observations on this subject, see above, Chapter 3, "Fallacies of Factual
Significance."

65. Hexter, Reappraisals in History, p. 213.

FALLACIES OF NARRATION

163

Hexter's electric imagery might be extended to describe other
dimensions of difficulty in this subject. There are not merely junction
boxes in narrative circuits, but transformers, capacitators, generators,
relays, breakers, and resistance coils — and there are radial circuits
and ring circuits and many other kinds as well. In common practice, the
problem of narrative connections is resolved by the use of an additional
device which is our business for the next chapter — causality.

CHAPTER VI

FALLACIES OF
CAUSATION

&*>

What is called Wisdom is concerned with primary causes.
— Aristotle

Der Glaube an den Kausalnexus ist der Aberglaube.
[Belief in the causal nexus is superstition.]
— Lud wig Wittgenstein

The causal principle is, in short, neither a panacea nor
a myth; it is a general hypothesis . . . having an approx-
imate validity in its appropriate domain.

— Mario Bunge

Today's historians tend, characteristically, to combine generalization,
narration, and causation in a single explanation model. But it is not
self-evident that problems of cause are properly a part of a historian's
work. Can any causal question be answered by an empirical method?
Can any existing model of causal explanation survive a rigorous logical
and linguistic analysis? These doubtful issues have occasioned a large
literature, but little agreement.

There was a time when nearly everybody believed, as some still
do, that the task of an historian was "to establish facts and to marshal
them in a sequence of cause and effect." 1 But early in the twentieth
century the idea of cause was widely condemned by scientists such as
Mach and Helmholz, by historians such as Beard and Teggart, by
novelists such as Kafka and Sartre, and by philosophers such as Witt-
genstein and Russell. The latter wrote, for example, "The law of
causality, I believe, like much that passes muster among philosophers,

1. Fritz Fischer, Germany's Aims in the First World War (New York, 1967), p. xxi.

FALLACIES OF CAUSATION

165

is a relic of a bygone age, surviving, like the monarchy, only because
it is erroneously supposed to do no harm." 2 Another philosopher has
argued that "causality may have no more reality than a dragon or a
mermaid." 3

Recently, however, causation has regained respectability and more
refinement than it ever possessed before, in discussions by philosophers,
physicists, and social scientists. But historians are, as usual, a genera-
tion behind. A few continue to condemn the whole idea. Professor
Howard Zinn has written,

I will not tangle with cause, because once you acknowledge cause as the
core of a problem, you have built something into it that not only baffles
people, but, worse, immobilizes them. Causation is not merely complex —
it is a problem impossible of solution according to some of the new [sic]
philosophers. Perhaps it is one of those metaphysical conundrums created by
our own disposition to set verbal obstacles between ourselves and reality. 4

Other historians have responded in a different way. They have
eliminated the word "cause" from their vocabulary, but they have
continued to construct cryptocausal interpretations. They have camou-
flaged causation behind words such as "influences," "impulses," "ele-
ments," "master symptoms," "prodromes," "mainsprings," "roots,"
"bases," "foundations," "undercurrents," "fountainheads," "fertilizing
factors," etc. All the synonyms in all the thesauri have been exhausted
by historians, in their efforts to avoid an explanation form they dis-
trust but have not been able to discard.

The most common semantical subterfuge is "factor." G. R. Elton
has observed that

The concept of cause being under a cloud (because it used to be employed
too crudely) historians have for a time been hunting for a substitute.
Wherever one turns in history today, one runs head-on into factors.
There are no longer any causes of the Reformation; instead there are
factors that made it possible. This is to go from the tolerably dubious to
the quite abominable. A cause is something real: people do things in order
to get results. A factor— outside mathematics, and trading stations and
Scottish estates — is a meaningless piece of tired jargon. Events are not
the product of simple causes but of complex situations in which a variety
of people and circumstances participates, but this does not mean that
they are produced by factors. A word to be forgotten. 5

2. Bertrand Russell, "On the Notion of Cause," Mysticism and Logic (London, 1958),
pp. 180-208. This work was first published in 1917.

3. William P. Montague, The Ways of Knowing, or The Methods of Philosophy (New
York, 1958), p. 199.

4. Howard Zinn, The Southern Mystique (New York, 1964), p. 7: see also the state-
ment by Charles Beard and Alfred Vagts in the Social Science Research Council's report,
Theory and Practice in Historical Study, Bulletin 54 (1946), pp. 136-37.

5. G. R. Elton, The Practice of History (New York, 1967), p. 101.

166

EXPLANATION

The problem is more than merely semantical. Historians have re-
cently responded to the problem of cause in the way that American
statesmen have reacted to the problem of China. Formal nonrecognition
masks a functional obsession. The results in both cases are some very
odd goings-on, which would be wonderfully entertaining, if their effects
were not so destructive.

There are, of course, many modes of historical explanation which
are noncausal in nature. But I have never read an extended historical
interpretation which does not include causal statements, or cryptocausal
statements, in at least a peripheral way. Whether or not historians ought
to do causal history, the plain fact is that almost all of them do it, and
some do it better than others. Someday soon we shall probably see the
refinement of a genuine and consistent noncausal history. But even then,
causal interpretations will probably coexist with other forms, because
causal relationships are something real in the world. Causation is an
idea which is generally used because it is generally useful. In some of
its renderings, it may be able to coexist with logic and empiricism. In
others, it is clearly fallacious. The object of this chapter is to identify a
few of the most common fallacies of causal analysis and then to con-
clude with a few affirmative remarks.

The fallacy of post hoc, propter hoc is the mistaken idea that if
event B happened after event A, it happened because of event A. An
example is provided by a female passenger on board the Italian liner
Andrea Doria. On the fatal night of Doria's collision with the Swedish
ship Gripsholm, off Nantucket in 1956, the lady retired to her cabin
and flicked a light switch. Suddenly there was a great crash, and grind-
ing metal, and passengers and crew ran screaming through the passage-
ways. The lady burst from her cabin and explained to the first person
in sight that she must have set the ship's emergency brake!"

There are many post hoc fallacies in historical scholarship. Con-
sider, for instance, one major interpretative problem — the defeat of
the Spanish Armada in 1588. Some scholars have suggested that this
event caused the decline of the Spanish empire and the rise of the Brit-
ish. But Garrett Mattingly replies that "it is hard to see why they think
so. By 1603, Spain had not lost to the English a single overseas outpost,
while the English colonization of Virginia had been postponed for the
duration." Others have argued that the defeat of the Armada transferred
control of the seas from the Spanish to the British. Professor Mattingly
replies that before 1588 "English sea power in the Atlantic had usually

6. Alvin Moscow, Collision Course (New York, 1959), p. 85.

FALLACIES OF CAUSATION

167

been superior to the combined strengths of Castile and Portugal, and
so it continued to be, but after 1588 the margin of superiority diminished.
The defeat of the Armada was not so much the end as the beginning of
the Spanish navy."

Still others have attributed to the defeat of the Armada the dis-
location of the Spanish economy by the disruption of communications
with America. Mattingly: "In fact, more American treasure reached
Spain in the years between 1588 and 1603 than in any other fifteen
years in Spanish history." Others still have argued that the Armada
"led to the great explosion of literary genius which marked the last
fifteen years of Elizabeth's reign." Mattingly: "The assertion of a causal
connection between the defeat of the Armada and the flowering of
Elizabethan drama, is hard to refute; even harder, except by the method
of post hoc, propter hoc, to prove. There is no link in England between
the Armada campaign and any literary work as clear as one [Don
Quixote] we can find in Spain."

Mattingly himself believed that there was something more to the
thesis that the Armada "decided that religious unity was not to be re-
imposed by force on the heirs of medieval Christendom." But this as-
sertion, moderate though it is in Mattingly's book, can be dealt with
and dismissed by the method of pro hoc. The permanence of the frag-
mentation of Christendom was surely clear before the Armada sailed.
If there were many Catholics or Protestants who remained blind to that
fact, the scales would not fall from their eyes for many years after the
Armada. A better argument can be made for the causal role of the
Armada in the growth of English nationalism, but even this would have
to be qualified both by stirrings of national spirit before 1588 and by the
persistence of subnational and supernational attachments in England
after that date.

In short, it appears that the defeat of the Armada, mighty and
melodramatic as it was, may have been remarkably barren of result.
Its defeat may have caused very little, except the disruption of the
Spanish strategy that sent it on its way. That judgment is sure to violate
the patriotic instincts of every Englishman and the aesthetic sensibili-
ties of us all. A big event must have big results, we think. But this is
the fallacy of identity, which we shall come to directly. 7

The fallacy of cum hoc, propter hoc mistakes correlation for
cause. Every textbook on statistics warns solemnly against this fallacy. 8

7. Garrett Mattingly, The Armada (Boston, 1959), pp. 397-402.

8. Frederick E. Croxton, Dudley J. Cowden, and Sidney Klein, Applied General

168

EXPLANATION

And, indeed, some very silly errors have flowed from it. Robert Garland,
the promoter of daylight-saving time, seriously suggested that boys who
lived in daylight time zones had bigger feet than boys in other areas,
and moreover that "bigger feet make better men." He concluded there-
fore that daylight time caused better men. 9

Another example is supplied by a statistician who reports that
there is a high positive correlation between the number of storks' nests
and the number of human births in various parts of northern Europe.
The apparent explanation is not that storks bring babies, but that popu-
lation increase correlates with an increase in building construction, and
that more buildings mean more places for storks to nest. 10

After the assassination of President Kennedy, many commentators
solemnly observed that since 1840, every President elected in a year
ending in has died in office — William Henry Harrison (1840), Abra-
ham Lincoln (1860), James A. Garfield (1880), William McKinley
(1900), Warren G. Harding (1920), Franklin D. Roosevelt (1940),
and John F. Kennedy (1960). A few lunatic numerologists were quick
to suggest a causal relationship.

There are many other examples in historiography. The Italian
economic historian Amintore Fanfani seriously argued that capitalists
tended to have longer heads than other people, and that therefore long-
headed people had better heads for business than shortheaded people. 11
More recently, Lee Benson has attempted to answer questions such as
"What was the cause of the Republican victory in 1860?" or "Was the
candidacy of James G. Blaine the cause of the Republican defeat in
1884?" strictly by means of correlation. 12

Correlation by itself can never establish a cause. It can disestab-
lish one — for there can never be a regularistic causal relationship without
correlation. But there can often be a regular correlation without a cause.
If X and Y occur together with perfect regularity, X may be the cause of
Y, or Y may be the cause of X, or Z may be the cause of X and Y, or
there may merely be a coincidence. Near-perfect correlations exist be-
tween the death rate in Hyderabad, India, from 1911 to 1919, and

Statistics, 3d ed (Englewood Cliffs, N.J., 1967), pp. 9, 406-7; W. Allen Wallis and
Harry V. Roberts, Statistics, A New Approach (New York, 1956), pp. 527, 529, 544;
A. L. O'Toole, Elementary Practical Statistics (New York, 1964), p. 243.

9. O'Toole, Elementary Practical Statistics, p. 243.

10. Wallis and Roberts, Statistics, p. 79.

11. Amintore Fanfani, Catholicism, Protestantism, and Capitalism (New York, 1938).

12. Lee Benson, "Research Problems in American Political Historiography," in Mirra
Komarovsky, ed., Common Frontiers of the Social Sciences (Glencoe, 111., 1957),
pp. 113-81.

FALLACIES OF CAUSATION

169

variations in the membership of the International Association of Machin-
ists during the same period. Nobody seriously believes that there is
anything more than a coincidence in that odd and insignificant fact. 13
To establish the regularistic causal proposition that X caused Y,
three things must be demonstrated. First, there must be a correlation
between X and Y. Second, there must be a proper temporal relationship
in their occurrence. Xi must occur before Yi. Third, there must be at
least a presumptive agency which connects them. There is solid evidence
that drunken driving causes accidents. A correlation has been observed
beyond all doubt; the proper temporal relationship exists, and there is
a chemical and physiological agency which is well understood. There
is pretty good evidence that cigarette smoking causes lung cancer. The
correlation is high, the temporal relationship is correct, and there is a
reasonable presumption of an agency in inhalation of tobacco smoke
and the deposit of carcinogens in the lung. The precise nature of that
agency is not as clearly understood as in the case of alcohol and acci-
dents, but the fact of its existence seems clear beyond a reasonable
doubt.

Sometimes the question of causality can be by-passed by an in-
vestigator. In my opinion, whenever it can be avoided, it should be.

In statistical work [one scholar has written] it is not necessary to prove that
there are cause-and-effect relationships in every situation. Often, the estab-
lishment of the fact that there exists an association, and the measurement of
its degree or intensity, are all that the investigator needs for the purposes of
estimation and prediction. The statistician often is more interested in what
he can accomplish by using observed associations or relationships than he is
in assigning cause-and-effect or other explanations to them. 14

This is surely true, and important. And yet, if an observer wishes
to get beyond estimation and prediction to the problem of control, or
to the problem of super- or extraregularistic explanation, then he must
address himself to the problem of causality, however irritating it may be.
And correlation, in conclusion, is only a necessary part of regularistic
causal explanation. It is not sufficient, in itself, to resolve such issues.

The fallacy of pro hoc, propter hoc consists in putting the
effect before the cause. It is an embarrassing error, which violates one
of the few simple and self-evident rules of causality: if event E happened
before event C, it cannot have happened because of it. In history, there

13. Harold A. Larrabee, Reliable Knowledge (Boston, 1954), p. 368.

14. OToole, Elementary Practical Statistics, p. 243.

170

EXPLANATION

are no prior effects and no instantaneous causes. The effect must always
follow the cause, and it must be separated by a temporal interval, be it
ever so slight. 15

The fallacy of pro hoc is the sort of blunder which always leaves its
perpetrator feeling a little queasy after it is pointed out to him. But its
absurdity is more apparent in a general description than in particular
instances. It has, indeed, been committed by two of the most able aca-
demic historians who are writing in America today: C. Vann Wood-
ward and David M. Potter.

Woodward made this mistake in an essay called The Age of Re-
interpretation, which first appeared as an article in The American His-
torical Review in October, 1960, and subsequently as a separate imprint
by the American Historical Association. In an argument as compre-
hensive as Frederick Jackson Turner's interpretation of the significance
of free land in American history, Woodward suggested that there was a
comparable importance in the existence of an alleged "free security"
for the development of American culture. Separated from its enemies
by three oceans, and protected by the ships of the Royal Navy, the
American Republic grew and prospered, Woodward believed, without
having to pay for a military establishment, such as less fortunate nations
required. It never had to pay a cost which was as heavy in political
and psychological and social and cultural terms as it was in dollars.

From this supposed cause, Woodward reasoned to many broad
effects: to the pace of American economic development, the prevalence
of light taxes and limited government, the indiscipline of American
youth, and other national characteristics which Woodward collectively
called the "sunnier side of the national disposition — the sanguine tem-
perament, the faith in the future, what H. G. Wells once called our
'optimistic fatalism.' "

It is a splendid argument. But unfortunately, the author is forced
to make one concession which is both empirically necessary and logically
fatal to his thesis. He concedes that before 1815 (or thereabouts),
there was little "free security" for Anglo-America. Our oceanic moat
was, in the first two centuries of our history, a great common, as Captain
Mahan once called it, which did not protect the colonies but exposed
them to their enemies. And there were enemies aplenty — formidable
Indian tribes, the Dutch, the French, the Spanish, and after 1775 the
British themselves. The southern colonies felt a continuous threat of

15. It is possible, of course, to have a cause C, and an effect E v and a C 2 and an E 2 ,
which happened precisely in that temporal order. C 2 thus happened after £,, and C l
and C 2 may be indistinguishable in all respects — all except one. C 1 caused E v and C 2
did not, and could not.

FALLACIES OF CAUSATION

171

domestic insurrection by its black labor force. And all the colonies
were menaced by privateers and pirate fleets which were sufficiently
powerful to lay siege to colonial capitals.

From the beginnings of settlement to 1815, there was a war in
every American generation, and some of these wars were cruel and
bloody. It is unlikely that a civilized society, anywhere in the world,
has ever survived losses in proportion to those Virginia experienced
in its first half century. And the deaths which New England suffered
during King Philip's War were greater in proportion to its population
than those which Russia or Germany sustained in World War II. The
Anglo-American population, of course, was very much smaller than
these great nations, but the social impact must have been comparable. 18

Anglo-America fought its earliest and bloodiest wars without
much help from the mother country, except in the French and Indian
War. It was forced to conquer its own peace, at a heavy cost in blood
and treasure. Something near to universal military service was forced
upon it, and military expenditures which it could ill afford. And yet,
in the midst of many tribulations, the phenomena which Woodward
interprets as the fruit of free security developed swiftly: freedom and
individuality, limited government, rapid economic growth, and all
the rest. The "sunnier side" of the national disposition is perfectly
personified by Benjamin Franklin, who lived much of his life in a
world of war. It is precisely described by Crevecoeur, who suffered
severely in the Revolution.

In short, Woodward, who has worked mostly in the period since
1876, made a mistake which is all too common among students of the
recent past. He forgot the first half of American history — the period from
the Virginia massacre to the battle of New Orleans, which is longer
than the interval between the battle of New Orleans and the siege of
Khesanh. Woodward's oversight might be politely described as dysfunc-
tional to his thesis, for it was the first half of American history which
marked the emergence of his effect, and the second half which included
the appearance of the alleged cause."

16. The casualties of Germany and Russia in World War II are estimated at one in
twenty-five and one in twenty-two of 1940 population. New England's losses in King
Philip's War were at least one in twenty. See Douglas E. Leach, Flintlock and Toma-
hawk (New York, 1966), pp. 10, 243.

17. Woodward could conceivably salvage something from the wreckage by making two
major alterations in his argument. He could claim that there was a certain measure
of "free security" in England before the colonization of North America and that the
culture of the colonists was affected by it. And he could argue that "free security," in
the period 1815-1949, had an intensifying or conservatizing effect with respect to the
many qualities which he attributes to it. But neither of these refinements appears in his
essay.

172

EXPLANATION

David Potter's fine book People of Plenty is also a splendid speci-
men of a pro hoc for our collection of logical curiosities. Potter makes
abundance the cause of much that is American. But America has not
always been possessed of abundance, or affluence. In the first two
centuries of our colonial and national existence, there were few great
American fortunes, nor was there much of a surplus of anything, except
rocks, trees, and trouble. The portraits of our ancestors have a lean
and hungry look. It might be said that they became American, and
then they became affluent. It is interesting to note that two centuries ago
many students of American society believed that its free institutions
were sustained by poverty rather than by wealth. From John Winthrop
to John Adams, Americans thanked God that they were not a wealthy
nation, for they believed that wealth spawned luxury and corruption
and despotism and all the other ugly things that the Old World allegedly
was and the New World wasn't — yet. The many national characteristics
Potter identifies came early in our history (except for things like the
advertising industry, which is my candidate for an un-American activ-
ity). Americanism came early; abundance came late. And a good
many Americanisms may be surviving in spite of our abundance — not
because of it.

$»» The reductive fallacy reduces complexity to simplicity, or
diversity to uniformity, in causal explanations. It exists in several
common forms, none of which can be entirely avoided in any historical
interpretation. As long as historians tell selected truths, their causal
models must be reductive in some degree. But some causal models are
more reductive than others. When a causal model is reductive in such
a degree, or in such a way, that the resultant distortion is dysfunctional
to the resolution of the causal problem at hand, then the reductive
fallacy is committed.

One common form of the reductive fallacy is the confusion of
necessary with sufficient cause — the confusion of a causal component
without which an effect will not occur, with all the other causal com-
ponents which are required to make it occur. This sort of error appears in
causal explanations which are constructed like a single chain and
stretched taut across a vast chasm of complexity. The classic example is
the legendary battle that was lost for the want of a horseshoe nail: for
the want of a nail the shoe was lost, for the want of a shoe the horse
was lost, for the want of a horse the rider was lost, for the want of a
rider the message was lost, for the want of a message the regiment was
lost, and for the want of a regiment the battle was lost.

FALLACIES OF CAUSATION

173

This exemplary anecdote has probably been told to every quarter-
master in the Western world — no doubt with a salutary effect upon the
supply of military horseshoe nails. I am told that our army still possesses
a considerable quantity of them. But it has not helped our sense of
history.

In the late summer of 1862, for example, the Confederate army
of Northern Virginia suddenly crossed the Potomac River and marched
north into Union territory, threatening Baltimore, and Washington, and
the rich farming country of south-central Pennsylvania. No Union
commander could be certain of Robert E. Lee's intentions, which were
in fact to concentrate his army near a small Maryland town and to
march into Pennsylvania. Lee so informed his lieutenants in a document
well known to Civil War buffs as Special Orders no. 191, copies of
which were made for all Confederate division commanders. By chance
two copies reached General D. H. Hill. One copy was carefully preserved;
the other was used by a staff officer to preserve his cigars against the
dews and damps of a September morning in Maryland. Somehow the
packet — cigars, special order, and all — slipped out of his pocket, and
into the hands of the Union general George B. McClellan. A few
Union special orders were promptly issued, and there was a fight
which the North called Antietam and the South called Sharpsburg and
many a weary infantryman on both sides must have known by yet
another name. It was the bloodiest day of the war, and a black one
for Confederate arms. When it was over, General Lee was forced to
retreat into Virginia. It is often said that Antietam was the decisive
battle of the war. Many historians believe that it ended all chance of
European intervention on the side of the Confederacy. Some are also
of the opinion that this victory permitted Abraham Lincoln to gain a
critical measure of control over his domestic opposition. Moreover, a
few days after the engagement, Lincoln issued his preliminary Eman-
cipation Proclamation.

Are we to conclude from this story that the cause of Northern victory
in the Civil War was the loss of Special Orders no. 191? The answer
depends upon the causal model which is at hand. There is, I think, no
prima facie case against the validity of such a causal interpretation, if it
is clearly understood that everything depends upon the acceptance of
a contingent-series model of causality, and if the question at hand can
be fairly and fully met by such an explanation. But there are many
other kinds of causal models (at the end of this chapter an attempt
will be made to identify them), and the questions which many historians
choose to ask today are not of the sort which can be satisfied by a
"want-of-a-nail" explanation. The reductive fallacy might therefore be

174

EXPLANATION

redefined as the asking of one kind of causal question, and the answering
of it with another and less comprehensive kind of causal explanation.

Other examples of reduction are undeniably fallacious in many of
their applications. There is the classical case of Actium, Antony, and
Cleopatra's nose. Why, one wonders, is Cleo's nose singled out for
special attention? Surely other anatomical parts were more important to
a red-blooded Roman.

Another famous example is supplied by Winston Churchill. In
1920, King Alexander of Greece died of blood poisoning, having
been bitten by a pet monkey. This event was followed by a plebiscite,
and a new king, and a bloody war with the Turks. Churchill wrote,
"A quarter of a million persons died of that monkey's bite." 18

There are endless other instances — Bajazet's gout, which allegedly
interrupted his ambition to feed his horse upon the altar of St. Peter
and prompted Gibbon to comment that "the disorders of the moral,
are sometimes corrected by those of the physical world; and an acrimon-
ious humor falling on a single fibre of one man may prevent or suspend
the misery of nations." 19 One thinks also of the Duchess of Marlborough's
gloves, and the Holy Roman Emperor's mushrooms, and Robert the
Bruce and the spider, and an endless miscellany of other reductive
curiosities.

Whole schools of historiography are grounded in the fallacy of
reduction. One of them is Marxist Leninism, which in Leonard Krieger's
phrase "is reminiscent of a Rube Goldberg mechanism in which a
fat nobleman opens the door at one end of a corridor and sets off a
chain reaction which explodes Nicolai Lenin out the door at the other
end." 20 Similar errors appear in psychoanalytical history, which would in
some of its forms reduce the cause of the reformation to Martin Luther's
toilet training. Good Marxists and good psychoanalytic historians rec-
ognize this form of error and repudiate it, but as long as they hold to their
various monisms, their explicit rejection of reduction is contradicted
by the implicitly reductive nature of their interpretations.

The most stubborn and dangerous forms of reduction appear not
in the works of historians of any persuasion but in the ravings of
various ideological true believers, for whom history is "a thin thread of
evidence taken from selected literary works."" Everybody, I suspect,
has had some familiarity with such people, who carry their archives in

18. The World Crisis: The Aftermath (London, 1929), p. 386.

19. Gibbon, The Decline and Fall of the Roman Empire, ed. J. B. Bury, 3 vols. (New
York, 1946), 3, chap. 64, p. 2236.

20. Leonard Krieger, "A View from the Farther Shore," Comparative Studies in Society
and History 5 (1963): 269.

21. George Thayer, The British Political Fringe (London, 1965), p. 31.

FALLACIES OF CAUSATION

175

their pockets — a few dog-eared clippings from periodicals and other
people's books, snippets and shreds of evidence which they are likely
to drag out on any argumentative occasion.

A curious backhanded twist upon the reductive fallacy occurs in
a recent book by William and Paul Paddock, ominously titled Famine —
1975! (Boston, 1967). These authors attempt to demonstrate the
inevitability of this disaster, by taking up one by one the various devices
and tactics which have been proposed to forestall it. They consider
the potential impact of various population limiting devices — IUD, the
pill, sterilization, and others. And they separately examine the probable
result of rising literacy rates and living standards, and various agricul-
tural innovations — synthetic proteins, vitamins, incaparina, hydropon-
ics, desalinization, the expansion of fisheries, the extension of agricultural
research, the increased use of fertilizers, irrigation systems, land reform,
government bounties and controls, private enterprise, and finally the
"panacea of the unknown panacea." Each of these panaceas is separately
rejected as insufficient to resolve the problem. The authors conclude
that the problem is therefore insoluble. But they do not consider the
effect of all of these measures, simultaneously applied — as they are
likely to be. In other words, they study the possible effects of these
measures reductively without considering their collective, interactive
effect. 22

$»> The fallacy of indiscriminate pluralism is the converse of
the reductive fallacy. It appears in causal explanations where the number
of causal components is not defined, or their relative weight is not
determined, or commonly both. The resultant explanation, for all its
apparent sophistication and thoroughness, is literal nonsense. It is
also self-contradictory, for an indiscriminate pluralism is not really a
pluralism at all, but a perverse kind of monistic unity comparable to
William James's idea of an infant's idea of the universe — "one big
blooming buzzing Confusion." 23

Indiscriminate pluralism is an occupational hazard of academic
historians, who are taught to tell comprehensive truths. It is particularly
powerful in the present generation, when all monisms are under the
ban, and a "single-factor" thesis rarely appears without an extended
prefatory apology. One hardly ever sees a contemporary reference to
the cause of an event, but often to a multiplicity of "causes," "factors,"
"elements," "origins," "influences," "impulses," "stimuli," etc. As

22. Pp. 62-97.

23. William James, Psychology (Cleveland, 1948), p. 16.

176

EXPLANATION

pluralism becomes more popular, indiscriminate pluralism becomes
more prevalent. It is understandably common among undergraduates and
graduate students who are more than a little bewildered by the mass of
competing interpretations which they are required to integrate into a
coherent pattern. Every generation of history graduate students confronts
twice as many books as the generation which preceded it. The number
of titles has doubled every twenty years (roughly) since the invention
of the printing press, and the number of monographs and learned
journals has recently been rising even faster.

An apprentice historian also faces a great and growing eclecticism
of method, technique, style, subject, and interpretation among the master
craftsmen whom he must satisfy before he is admitted to the guild. Indis-
criminate pluralism thus becomes functional to his professional purposes,
at the same time that it is dysfunctional and deleterious to his inter-
pretations. An example, possibly apocryphal, is a history graduate
student in a great American University, who is said to have equipped
himself for his General Examination with a set of cards, each of which
carried a single word — "political," "economic," "constitutional," "re-
ligious," "military," "intellectual," "educational," "diplomatic," "demo-
graphic," "cultural," "social," and "miscellaneous." He is reputed to
have responded to each question which was put to him by consulting
his cards and enumerating "factors" in every category, giving equal
time to each of them, and he passed with distinction!

But the fallacy of indiscriminate pluralism is not confined to
students. Many skilled and mature historians have allowed it to creep
into their work. One instance is an excellent book on the coming of the
American Revolution by Bemhard Knollenberg. The author is not an
academic historian but a lawyer by profession. He is, however, a fine
scholar, and a worthy successor to the tradition of the great nineteenth-
century gentleman-historians — an amateur in all the best senses of that
battered word, and none of the worst.

Knollenberg's book, the first volume of a projected series, is an
account of British imperial policy and American colonial politics from
1759 to 1766. He rejects several monist interpretations which have
located the cause of the war in economics, or the alleged closing of the
frontier, or politics. Instead, he offers a long list of irritants, which cover
every category in our graduate student's cards, including "miscellaneous."
But he does not carefully weigh one against the other in an integrated
and refined interpretation. The result is more useful in its various parts
than in the whole, which is shapeless and diffuse."

24. Bemhard Knollenberg, Origin of the American Revolution: 1759-1766, rev. ed.
(New York, 1961).

FALLACIES OF CAUSATION

111

Indiscriminate pluralism is often implicit in works of narrative
synthesis. One example, in a beautiful piece of narrative craftsmanship,
is Johan Huizinga's The Waning of the Middle Ages, in which things
have, in the author's words, "all sorts of causes of a general nature —
political, economic, ethnographic," etc. etc. 2 '"'

The fallacy of identity is the assumption that a cause must
somehow resemble its effect. It is related to the idea which explicitly
underlay many folk remedies, the so-called doctrine of signatures,
which was a belief that "every natural substance which possesses any
medicinal virtue indicates by an obvious and well-marked external
character the disease for which it is a remedy, or the object for which
it should be employed." 26 Examples are the idea that turmeric is a cure
for jaundice, or bloodstone for bleeding.

When the fallacy of identity appears in modern historical scholar-
ship, it is apt to be implicit, rather than explicit, but its effects are no
less troublesome for that fact. One example is the historiography of
early inhabitants of Scotland. The Picts constructed brochs and souter-
rains which are small, dark, and mysterious. From this, some have
concluded that the Picts themselves were small, dark, and mysterious —
an inference which is described as "romance" by a good historian. 27

A more common form of the fallacy of identity is the idea that big
effects must have big causes, or that big events must have big con-
sequences. We have already met this assumption in the historiography
of the Spanish Armada. What was the cause of the collapse of the
Manchu dynasty? One historian has written, "So swift a decline, so
unexpected a reversal of fortune must have some deep-seated cause." 28
But must it?

An even more common form of the fallacy of identity appears
often in what J. H. Hexter has called Tunnel History. There is a
tendency in topical works to assume that economic effects have primarily
economic causes, and that the origins of a religious phenomenon are
necessarily religious, and that the great happenings in the history of
education are to be explained primarily in terms of earlier great happen-
ings in the history of education. The narrowness of such thinking
is to be explained in a variety of ways, which we shall consider in a

25. Johan Huizinga, The Waning of the Middle Ages (Garden City, N.Y., 1954).
p. 20.

26. John Stuart Mill, A System of Logic, 9th ed. (London, 1875), p. 502.

27. J. D. Mackie, A History of Scotland (Baltimore, 1966), p. 27.

28. C. P. Fitzgerald, The Birth of Communist China (Baltimore, 1964), p. 30.

178 EXPLANATION

later chapter. But part of the explanation, perhaps, is the hold which
the doctrine of signatures still has upon our thought.

The fallacy of absolute priority assumes that there must be
an absolute first term in any causal series, and that if event Ay causes
event Bi, the same cannot be true of Si and A2. 2 *

There are many problems of historical causality in which an inter-
active relationship between A„ and B„ makes better sense than any
other. Did Protestantism cause capitalism to develop swiftly in America
and Europe? Or did capitalism cause Protestantism to expand? Histor-
ians have been arguing about this problem, sometimes in these terms,
for three generations. Surely, there was an interaction between these two
great movements. It is impossible to say which came first, for there
were proto-Protestantisms and proto-capitalisms which can be traced
back to the Book of Genesis.

American historians seem equally determined to find a first cause
in another problem, which is also three historiographical generations old.
Did the inferiority of American Negroes cause anti-Negro prejudice or
did prejudice cause inferiority? Arguments have been advanced for
both of these positions. But Gunnar Myrdal has argued more persuasively
that there was a vicious circle, in which prejudice caused inferiority and
inferiority caused prejudice. There was intense anti-Negro feeling
in Anglo-America from the very beginning, but also the nature of an
African Negro's cultural heritage, the nature of Anglo-American cul-
ture, and the nature of the acculturative process were such that a Negro
was in a position and in a condition of cultural (but not racial) inferi-
ority from the very start.

$»» The fallacy of the mechanistic cause, so-called by R. M.
Maclver,

treats the various components of a social situation, or of any organized
system, as though they were detachable, isolable, homogeneous, independently
operative, and therefore susceptible of being added to or subtiacted from the
causal complex, increasing or decreasing the result by that amount. But even
a slight acquaintance with the mechanism itself should teach us to avoid this
fallacy. We find writers who tell us that juvenile delinquency is due so much
to this factor and so much to that and so much to this other. But no mech-
anic would make the mistake of saying that the carburetor contributed so

29. Morris R. Cohen and Ernest Nagel, Introduction to Logic and Scientific Method
(New York, 1934), p. 385.

FALLACIES OF CAUSATION

179

much and the ignition system so much and the gasoline so much to the
speed of the car. 30

Maclver's analogy is unfortunate, but his point is sound, in its
relevance to historical explanations. There is, I think, an unhappy
tendency for historians to break down the components of a causal
complex and to analyze them separately, and even to assess separately
their causal "influence," independent of other elements with which
they interact. Almost any major causal problem can serve as an
example — the cause of the American Civil War or the cause of World
War I. Historians have formed the habit of speaking of the causes of
these and other events — a usage which is technically incorrect, and also
misleading in its implication that the various "causes" can be individually
analyzed and assessed, and that somehow if one of the "causes" were
removed, the effect would be diminished in the proportion of that causal
component to other causal components. Imagine that an effect E was
caused by A, B, C, and D. If all of these four causal components were
necessary to that effect, then the removal of any one of them would not
diminish E by one-fourth. Its absence would make E impossible. On
the other hand, it is easy to imagine that A, B, C, and D, though not
individually necessary to E, nevertheless interacted in a geometrical
ratio. If there were only A, then E would be of a magnitude 1. If there
were only A and B, then the effect would be not 2 but 2 2 , or an E of
magnitude 4. A, B, and C would produce an E of 9, and all four causal
components, an £ of 16. This is an involved way of saying that a causal
complex is something other than the sum of its parts.

Three examples of the mechanistic fallacy are provided by Morton
White. One is from the work of a great classicist, J. B. Bury, who
eliminated depopulation, the Christian religion, and the fiscal system
as causes of the dismemberment of the Roman empire. "If these or
any of them were responsible for [the Empire's] dismemberment by the
barbarians in the West, it may be asked how it was that in the East,
where the same causes operated, the Empire survived much longer intact
and united." 31 But this is a mistake. The three causal elements which
Bury rejects may have interacted with each other, and with still other
elements, in such a way as to produce very different results in the West
and the East.

Another example is the criticism of Herbert Baxter Adams's
explanation of the cause of America's early democratization in terms

30. R. M. Maclver, Social Causation, rev. ed. (New York, 1964), p. 94.

31. J. B. Bury, History of the Later Roman Empire, 2 vols. (New York, 1958) 1:
308-9; quoted in Morton White, The Foundations of Historical Knowledge (New
York, 1965), p. 65.

180

EXPLANATION

of the "Germanic tradition." Many critics of this thesis have replied that
the "Germanic tradition" must be present in Germany, which is not
notorious for its democracy. But, White observes, "these supposed
refutations of singular explanatory statements are faulty if they proceed
on the assumption that Herbert Baxter Adams, when he said that the
presence of the Germanic tradition was responsible for American dem-
ocratic institutions, necessarily implied that wherever the Germanic tra-
dition is present, democracy arises." 32 It is possible that such a cause as
this could produce very different results in different contexts.

A third example is from the great French medievalist, Marc Bloch.
White writes:

Even Marc Bloch, who was so much more concerned with the logic of his
discipline than most historians, seems to have been involved in a logical in-
consistency in his treatment of two distinct explanatory statements, just be-
cause he seems to attack one by fallaciously arguing that it implies a false
generalization but fails to use the same fallacious method of refutation on
another. In his Feudal Society he considers as an explanation of the cessation
of Scandinavian pillaging in the Middle Ages the fact that the Scandinavians
were converted to Christianity, and apparently rejects this explanation on
the ground that it implies the false generalization that no Christian people
would engage in pillaging. He says: "As we shall often have occasion to ob-
serve in the following pages, among the peoples of the West during the feudal
era there was apparently no difficulty in reconciling ardent faith in the Chris-
tian mysteries with a taste for violence and plunder, nay even with the most
conscious glorification of war." On the other hand, when Bloch tries to ex-
plain the beginning of the Scandinavian invasions, he is prepared to accept
as its explanation the fact that the Scandinavian countries were overpopulated
at the time, though he surely would deny that the people of every overpopu-
lated country invade in the manner of the medieval Scandinavians. What we
have here is a kind of double standard. 93

The fallacy of reason as cause mistakes a causal for a logical
order, or vice versa. It is a form of error for which the ambiguity of
the term "because" may be partly responsible. We use that word both
as a conjunction in a causal proposition and as a conjunction in a logical
proposition. It is easy, in practice, to confuse the two. Consider the
following statements:

1. Cromwell died because he caught intermittent fever.

2. Cromwell died because all men die, and Cromwell was a

man.

The first sentence supplies a causal explanation which is acceptable to

32. White, p. 64.

33. White, pp. 65-66; quoting Marc Bloch, Feudal Society (Chicago, 1961), pp. 35-38.

FALLACIES OF CAUSATION

181

historians but syllogistically invalid. 34 The second sentence is syllogisti-
cally valid but historically insignificant and causally meaningless. To
mistake it for a causal explanation is fallacious.

Sometimes the error runs in the opposite direction, and a causal
explanation is mistaken for a reasoned argument. This was the case
in the doctrines of historical relativism, which were fatally flawed by a
classic example of the fallacy of reason as cause. Relativism confuses
two very different problems — the problem of how knowledge is acquired
and the problem of the validity of that knowledge. The fact that
historical knowledge is itself historically caused by the situation of the
historian does not in any degree imply that it is false. Blind patriotism
may cause a Polish historian to assert that a German army invaded
Poland in 1939. That statement possesses a truth value which is inde-
pendent of its origins. A German historian might be similarly motivated
to insist that a Polish army invaded Germany in 1939. That statement,
whatever its cause, is false. The same logic equally applies to factual
statements of every magnitude, though its application becomes more
complex in a geometric ratio to the increase in size.

There are many other errors in relativism — errors which have
been discussed at considerable length by competent philosophers. 3r '

34. Some philosophers think that causal explanations of this sort are enthymenes —
abbreviated syllogisms, in which a premise is not made explicit. But this, I think, is
wrong. There could be an implied general premise here, but not one which will
sustain a syllogism. The implied premise might be something like, "Some men who
catch intermittent fever die." Or it might be "Most men (of great age, frail frame,
etc.) who catch intermittent fever die." But these statements cannot satisfy a rule of
syllogistic reasoning which requires that the middle term must be "distributed" at
least once — i.e., that it must refer to everything in its range of reference, and not
merely to some or to most things. Without a distributed middle term, a deductive
syllogism collapses. No statement of statistical regularity can be properly distributed,
and yet, no other kind of regularity statement is possible in historical writing.

But even a statement of statistical regularity is not, I think, what a historian
commonly has in mind in causal propositions of this sort. He means rather to say
that intermittent fever is a disease which can kill a man, and it did kill Cromwell.
That it can be fatal is something which he might learn from merely a single instance,
by an empirical understanding of the causal linkages, even though it may only have
killed one man and no others. It is not always necessary, in history, or the natural
sciences, or daily life, to have a causal explanation which includes either a general
statement or deductive certainty. All that is required in many cases is a demonstration
that the causal relationship is possible, and that it probably happened. I suspect that
analytical philosophers who have attempted to impose the so-called covering law model
upon this species of explanation are guilty of an inverse form of the fallacy of reason
as cause, in their confusion of two different forms of thought.

35. A good critique of Charles Beard's relativism appears in Arthur C. Danto,
Analytical Philosophy of History (Cambridge, 1965), chap. 6; and Mannheim's
"relationism" is refuted in Charles Frankel, The Case for Modern Man, 2d ed. (Boston,
1959), chap. 7. See also, Christopher Blake, "Can History Be Objective?" Mind 72
(1955): 61-78; Ernest Nagel, "Some Issues in the Logic of Historical Analysis,"

182

EXPLANATION

And there is also something which is true and important. It is surely
correct that no historian can know the totality of history as actuality.
And it is a plain fact that history is something which happens even to
historians. But the corrosive skepticism which relativism engendered
was both logically deficient and practically deleterious to the progress
of historical knowledge.

The fallacy of responsibility as cause confuses a problem of
ethics with a problem of agency in a way which falsifies both. It often
consists in merging two different questions and demanding a single
answer: "How did it happen?" and "Who is to blame?" This pernicious
practice is particularly common in attempts to explain disagreeable
jvents, which are mostly contemporary events. An economist ran into
it in a practical situation. Paul Samuelson writes,

In some traditions a responsible and free human agent has come to be re-
garded as the "cause" of anything. When Professor Jacob Viner and I served
on the advisory board to the Commission on Money and Credit, I was in-
terested in hearing him remark that there was good precedent in the fields
of jurisprudence and torts to lay any possible blame for postwar inflation
upon the Federal Reserve Board rather than on such factors as the backlog
of demand or level of public debt, since "they" were responsible agents whose
duty it was to prevent the evil. 36

The most glaring example in American historiography is the at-
tempt to explain the course of Reconstruction after the Civil War. The
question of cause is no sooner raised than it is transformed from "What
caused it?" to "Who is to blame?" The answers are numerous, but nearly
all of them seek to impose responsibility upon some human agent, who
has been specifically identified as Andrew Johnson, or Thaddeus Stevens,
or the Radical Republicans collectively, or intransigent white Southern-
ers, or Northern businessmen, or carpetbaggers, or scalawags, or drunken
Negro legislators.

But it is quite impossible to locate any individuals who were re-
sponsible in both a moral and a causal sense for what happened in this
painful chapter of our past. The cause of the failure of Reconstruction
race policy must surely be sought in general phenomena for which no

Scientific Monthly 74 (1952): 162-69; Jack W. Meiland, Scepticism and Historical
Knowledge (New York, 1965); and Cushing Strout, The Pragmatic Revolt in American
History (New Haven, 1958), p. 59. passim.

36. Paul A. Samuelson, "Some Notions on Causality and Teleology in Economics,"
in Daniel Lemer, ed., Cause and Effect (New York, 1965), p. 100 n.

FALLACIES OF CAUSATION

183

free and responsible human agent can be held to blame. First, there
was an intensity of Negrophobia in the nation, which resisted all cor-
rection. The real sovereigns of the United States, the white majority,
were unable to accept the fundamental proposition that Negroes were
people. Second, there was the prevalence of a tacit moral philosophy in
the United States which was very close to that of Jefferson and Adam
Smith — a philosophy which held that governmental power could not
help people who were unable to help themselves, without destroying
their moral fiber. Third, there was the ugly but undeniable fact of the
brutalization of black slaves, who were incompetent as citizens of a
free republic. None of these conditions was subject to modification by
any act of any leading statesman. To argue the proposition "Was
Andrew Johnson to blame, or Thaddeus Stevens?" is to manifest an em-
pirical limitation and perhaps a moral blindness as well. It is also logi-
cally indefensible.

What is a causal explanation? It is an attempt to explain the
occurrence of an event by reference to some of those antecedents which
rendered its occurrence probable. This definition is ambiguous as to the
number and nature of the antecedents. The ambiguity is deliberate, be-
cause historians do different things when they attempt to construct causal
explanations. I can find no a priori superiority in any one of these differ-
ent causalities. But it is important to recognize their differences, for
error has often resulted from confusing one kind of causal explana-
tion with another.

John Stuart Mill believed that "the real cause is the whole of these
antecedents, and we have, philosophically speaking, no right to give
the name of cause to any one, exclusively of others." 37 This assertion may
be true, but it is not very useful. For some small events, a historian may
hope to recover all of the antecedents which rendered the effect possible,
or probable, or necessary. But in most causal problems which historians
actually confront, selection is not merely desirable but inescapable.

Which antecedents should be selected? Philosophers and, implicitly,
historians have developed many different criteria. R. G. Collingwood
alone has suggested three of them. First,

that which is "caused" is an event or state of things standing to it in a one-to-
one relation of causal priority; i.e., a relation of such a kind that (a) if the
cause happens or exists the effect also must happen or exist, even if no further
conditions are fulfilled, (b) the effect cannot happen or exist unless the cause

37. Mill, A System of Logic, p. 214.

184

EXPLANATION

happens or exists, [and] (c) in some sense ... the cause is prior to the effect,
for without such priority there would be no telling which was which. 38

Collingwood calls this a "theoretical natural science" type of causal
explanation. But it is sometimes used in history, as when an economic
historian explains the fact of inflation by the quantity theory of money. 39
The regularistic relationship in this model is always of a statistical sort,
and a ceteris paribus clause must always be included. This requires an
amendment in Collingwood's definition. A historian, and indeed a
natural scientist, can never assert that an effect will always happen
but only that it will probably happen. The connection between cause and
effect is not necessary but probabilistic.

Collingwood identified a second kind of causal explanation, in
which "that which is 'caused' is an event in nature, and its 'cause' is an
event or state of things by producing or preventing which we can pro-
duce or prevent that whose cause it is said to be." He attributes this
kind of causal explanation to "practical science," but it is also found
in history. It is used, I think, by Eric McKitrick in his causal explanation
of the course of Reconstruction after the American Civil War. 40 This
model of causal explanation as controllable antecedents raises a problem
of reference — namely, controllable by whom? Collingwood's answer is
a relative one, in which every man becomes a historian, and the con-
trollable antecedents are those he can control:

A car skids while cornering at a certain point, strikes the kerb, and turns
turtle. From the car-driver's point of view the cause of the accident was
cornering too fast, and the lesson is that one must drive more carefully. From
the county surveyor's point of view the cause was a defect in the surface or
camber of the road, and the lesson is that greater care must be taken to make
roads skid-proof. From the motor-manufacturer's point of view the cause
was defective design in the car, and the lesson is that one must place the
centre of gravity lower. 41

In actual road accidents, however, a variant upon this causal cri-
terion is regrettably more common. People involved in accidents tend
to assume that the cause of their misfortune was an antecedent which
somebody else could control. The driver would blame the surveyor for
the condition of the road or the manufacturer for the quality of the

38. R. G. Collingwood, An Essay on Metaphysics (Oxford, 1940), pp. 285-86. All
quotations from Collingwood in this section are from this source, unless otherwise
attributed.

39. Earl J. Hamilton, American Treasure and the Price Revolution in Spain, 1501—
1560 (Cambridge, 1934), p. 301. Compare, however, J. H. Elliott, Imperial Spain,
pp. 189-94.

40. Eric L. McKitrick, Andrew Johnson and Reconstruction (Chicago, 1960), passim.

41. Collingwood, An Essay on Metaphysics, p. 304.

FALLACIES OF CAUSATION

185

car. The surveyor would condemn the driver, or the manufacturer. The
manufacturer would point to the surveyor, or the driver.

A historian uses neither of these variants, for he is not a partici-
pant and there is rarely an antecedent which he personally can control.
He conventionally chooses some or all of the principal participants and
identifies controllable antecedents by reference to them, as McKitrick
did. The same strategy is also used in daily practical decision making.

A third form of causal explanation is one which Collingwood be-
lieved to be the historical form, and one which he used himself. In "his-
torical causation," he wrote, "that which is 'caused' is the free and
deliberate act of a conscious and responsible agent, and causing him to
do it means affording him a motive for doing it." This, of course, applies
only to deliberate and free acts, which comprise but a small fraction of
events which historians actually attempt to explain. But within this limit,
it is a form of causal explanation which is widely used.

A fourth form of causal explanation has also been suggested — one
which consists in the identification of abnormal antecedents. 42 An ex-
ponent of this view, R. M. Maclver, explains that an abnormal phe-
nomenon is one which "intrudes on a relatively constant system or
disturbs a relatively constant equilibrium. . . . Man has set up within
or upon the order of nature a vast number of temporary operative sys-
tems, subject in their degree and after their kind to disturbances from
within and from without." It might appear that this position entails a
powerful metaphysical bias toward equilibrium as normal, and so it
does in Maclver's version of it. But it need not do so. A historian could
explain the phenomenon of political stability in terms of an abnormal
equilibrium between economic and demographic growth.

In any case, causal explanation is often understood, in ordinary
historiographical practice, to mean the identification of abnormal ante-
cedents. Pirenne uses it, in precisely the terms of Maclver's model, to
explain the cause of the disruption of the "Mediterranean commonwealth"
by the invasion of Islam: "Now, all of a sudden, the very lands where
civilization had been born were torn away; the Cult of the Prophet was
substituted for the Christian Faith, Moslem law for Roman law, the
Arab tongue for the Greek and Latin tongue." 43

Still other historians have assumed that a causal explanation is one
that identifies "underlying conditions" which were of such a nature that

42. Maclver, Social Causation, p. 186; H. L. A. Hart and A. M. Honor£, Causation
and the Law (Oxford, 1959), pp. 31-38.

43. Henri Pirenne, Medieval Cities (Garden City, N.Y., 1956), p. 16. For this
example, and much more on the problem of causation, I am obliged to White,
Foundations of Historical Knowledge, p. 116.

186

EXPLANATION

they rendered the effect probable. This is perhaps the most common
of all forms of causal explanation in historical scholarship. Indeed,
causal explanations that depart from this model are commonly con-
demned in professional journals, for superficiality. An earlier genera-
tion of historians tended to use a different causal model, which was
constructed of a series of contingent events, rather than underlying
states. 44 Finally, some historians seek the precipitant cause. Kenneth
Stampp did something like this in And the War Came.

These various attempts to resolve the problem of which causal
antecedents should be included in a causal explanation yield eight an-
swers:

1. All antecedents

2 Regularistic antecedents

3. Controllable antecedents

4. Rational and/or motivational antecedents

5. Abnormal antecedents

6. Structural antecedents

7. Contingent-series antecedents

8. Precipitant antecedents

There are undoubtedly other criteria, and subdivisions of these, and
also ways of combining some of them with others. But this analysis
may serve to make the major point, which is that there are many different
kinds of causal explanation, and that they have different requirements
and different uses. The specific kind of causal explanation a historian em-
ploys must be selected according to the nature of the effect to be ex-
plained and the nature of the object of the explanation. Every causal
explanation should be an explanation to some purpose. There is no such
thing as the cause, and no cause for all occasions.

Most of the trouble historians get themselves into in causal ex-
planation consists in asking one kind of causal question and seeking
another kind of causal answer. Or it consists in a stubborn determination
to locate the cause. And both of these problems are aggravated by the
unfortunate tendency of historians to hide their causal models from
everybody — including themselves.

44. An excellent discussion, with examples, appears in White, Foundations of Historical
Knowledge, pp. 133-47.

CHAPTER VII

FALLACIES OF
MOTIVATION

3*

Men are not machines. . . . They are men — a tautology
which is sometimes worth remembering.

— Gilbert Ryle

Motivational explanation might be understood as a special kind of causal
explanation in which the effect is an intelligent act and the cause is the
thought behind it. Or it might be conceived in noncausal terms, as a
paradigm of patterned behavior. My own preference runs to the second
of these propositions, but there is no need to argue their relative merits
here. The following observations are consistent with both.

Historians have often used motivational explanations in their work.
Almost always, they have used them badly. Problems of motive in
academic historiography tend to be hopelessly mired in a sort of simple-
minded moralizing which is equally objectionable from an ethical and
an empirical point of view. Lord Rosebery once remarked that what
the English people really wished to know about Napoleon was whether
he was a good man. 1 The same purpose often prevails among profes-
sional scholars who are unable to. distinguish motivational psychology
from moral philosophy, and even unwilling to admit that there can
be a distinction at all. Moreover, many scholars tend to find fiat, monis-
tic answers to complex motivational problems, which further falsifies
their interpretations. 2

1. Quoted in E. H. Carr, What Is History? (New York, 1962), p. 97.

2. See, e.g., C. A. Macartney, Hungary: A Short History (Chicago, 1962), in which

188

EXPLANATION

Secondly, the quality of motivational explanation in history has
been diminished by the rationalistic bias of historians. William L. Langer
complained of his colleagues in 1957 that "Almost without exception . . .
they have stuck to the approach and methods of historicism, restricting
themselves to recorded fact and to strictly rational motivation." 3 Langer
proposed that the historian's "next assignment" was a "deepening of
our historical understanding through exploitation of the concepts and
findings of modern psychology." But ten years after his address, few
historians have acted upon his advice. Many scholars, of course, have
made use of the classic mechanisms of psychoanalytic theory — compensa-
tion, repression, identification, projection, sublimation, displacement. It
is rare to find a historical biography which does not introduce these
ideas. But they are used superficially, crudely, and inconsequentially,
as rhetorical supplements to motivational explanations which are still
predominantly rationalist.

A minority of historians have attempted to deal with the problem
of motivation in a more satisfactory way, by applying Freudian theory
directly to their scholarship. As early as 1913, an American historian
tried his hand at a psychoanalytic interpretation of Martin Luther. 4
There have been many other projects of a similar nature in the past
half century, and a few have generated useful and constructive insights
into historical problems. 5

But these experiments have ended in failure more often than suc-
cess. They have commonly consisted either of Freudian raids upon his-
tory, or of historians' raids upon Freud. The results have ranged from
the highly dubious to the downright preposterous.

If we are wise, we might learn something from both the successes and

principal figures are categorized in a spirit suggestive of 1066 and All That. We are
solemnly informed that "Stephen II was almost entirely bad." Louis I, on the other
hand, was "a true paladin," and Ladislas I was "a true paladin and a gentle knight"
(pp. 20, 42).

3. William L. Langer, "The Next Assignment," The American Historical Review
63 (1958): 283-304.

4. Preserved Smith, "Luther's Early Development in the Light of Psychoanalysis,"
American Journal of Psychology 24 (1913): 360-77.

5. Alexander and Juliette George, Woodrow Wilson and Colonel House: A Personality
Study (New York, 1956); William B. Willcox, Portrait of a General: Sir Henry Clinton
in the War of Independence (New York, 1964); Fawn Brodie, Thaddeus Stevens
(New York, 1959); Erik Erikson, Young Man Luther, A Study in Psychoanalysis and
History (New York, 1962); E. R. Dodds, The Greeks and the Irrational (Berkeley,
1951); Lewis J. Edinger, Kurt Schumacher: A Study of Political Behavior (Stanford,
1965); Rudolph Binion, Frau Lou: Nietzsche's Wayward Disciple (Princeton, 1968);
Lewis Namier, "King George III: A Study of Personality," Personalities and Powers
(London, 1955), pp. 39-58; and despite criticism from the cognoscenti, Ernest Jones,
The Life and Work of Sigmund Freud, 3 vols. (New York, 1953-1957).

FALLACIES OF MOTIVATION

189

the failures of Freudian historiography. Maybe the latter derive in some
degree from limitations in psychoanalytic method, as well as from the
poverty of historiographical procedure. Perhaps the utility and relevance
and accuracy of Freudian theory is seriously diminished by five sub-
stantial failings. First, it is in its aboriginal condition narrowly culture-
bound. I think it is significant that the success of psychoanalytic history
has tended to vary inversely with the temporal and spatial and cultural
distance between the subject and Freud. Second, psychoanalytic theory
has, in effect, extrapolated normative patterns from a study of neurotic
behavior. It is, again, no accident that psychoanalytic history has been
successful in proportion to the degree of serious mental disturbance in
the subject. For an investigation of motivation in stable, integrated
personalities, its value has been nearly nil. Third, psychoanalytic theory
places an excessive interest in childhood. This imbalance is untenable
in itself, and also exceedingly inconvenient to historians who cannot
hope to find the kind of evidence which psychoanalytic interpretations
require. Fourth, a similar problem derives from the hypersexuality of
Freudian theory, which is inherently indefensible and inapplicable to
historians' problems. Finally, psychoanalytic theory has often been built
into a closed deterministic system. Its enthusiasts have insisted that his-
torians must take all or nothing. Many scholars have chosen the latter
alternative, as the lesser evil.

These thoughts, of course, are not new. They have often been
articulated in the past fifty years. Too often, they have been offered
as excuses for a total rejection of psychological insights of all kinds into
history. The argument here is something different. It is not for ante-
Freudian history, or anti-Freudian history, or sub-Freudian history, but
rather for post-Freudian or super-Freudian historical scholarship.

Historians might be able to make much more effective use of post-
Freudian psychoanalytic literature than they can of the work of Freud
himself. One historian of modem America 6 has already discovered a
special relevance in the neo-Freudian thought of Karen Homey, who
attempted to adjust psychoanalytic theory to the cultural conditions she
discovered in the United States.

Moreover, historians might have more to learn from recent work
in social psychology than in psychiatry. The problems of social psychol-
ogy are closer to those of history, and the methods are more compatible.
At the end of this chapter, I shall have some more specific things to
say about the relevance of work by Abraham Maslow and David C.
McClelland.

6. David M. Potter, People of Plenty (Chicago, 1954), pp. 34, 54-72, passim.

190

EXPLANATION

Finally, if historians have much to learn from psychology, they also
may have something to teach. Psychology and psychoanalysis have both
been diminished by a temporal provincialism of the sort which only
historical scholarship can correct. But first, historians must put their
own house in order. Let us proceed to some common fallacies in
motivational explanation, and then to a few proposals for their correc-
tion.

The pathetic fallacy is the ascription of animate behavior
to inanimate objects. In Arnold Toynbee's definition, it is "imaginatively
endowing inanimate objects with life." 7 Most commonly, the pathetic
fallacy takes the form of anthropomorphism and anthropopathism, in
which human form and human feelings are given to gods, groups, objects,
etc. 8 There are many examples of the pathetic fallacy in explicitly his-
torical writing — by conservatives who fear the beast of Bolshevism,
liberals who complain of the cunning of capitalism, intellectual historians
who speak of the mind of the Enlightenment, and institutional historians
who would have a bare-breasted Madame Liberty hitch up her skirts and
hurdle the barricades, as in Delacroix's famous version of the Revolution
of 1830.

It is often difficult to distinguish this fallacy from what may be
merely an overblown figure of speech. And it is equally problematical
to locate the point at which the behavior patterns of individuals can be
transferred to groups. Is there such a thing as a national character? For
better or for worse, many historians and social scientists believe that
there is, and some of them speak of it as if it were a person. Sometimes
the intention is merely metaphorical. But it is an exceedingly doubtful
and dangerous image to introduce into one's thought, for it has a way
of spreading swiftly out of control. Consider the case of Henry Steele
Commager's book, The American Mind. In his preface, the author
explicitly acknowledged the fictional quality of his title. But in his
first chapter the fiction was reified, as Commager began to sketch with
bold, quick strokes of his pen a portrait of "The American." Sixty million

7. Arnold J. Toynbee, A Study of History, 12 vols. (New York, 1935-61), 1:8; 12:45,
116.

8. Sometimes other forms of life become tile model, as when Theodore Roosevelt
wrote a speech in which he made the world into a menagerie, and compared all the
nations to specific animals such as monkeys, hyenas, and hippopotami — an interpretative
device which his wise friend Elihu Root discouraged, not for its logical deficiencies,
but simply for the sake of its diplomatic shortcomings. Root solemnly suggested that
Roosevelt should strike out all sentences which might bring war betweeen the United
States and other countries.

FALLACIES OF MOTIVATION

191

minds magically became one, and a long list of singular characteristics
was recited:

The American was incurably optimistic. ... He had little sense of the
past. . . . The American had spacious ideas, his imagination roamed a con-
tinent, and he was impatient with petty transactions. ... He preached the
gospel of hard work. . . . All this tended to give a quantitative cast to his
thinking. . . . Theories and speculations disturbed the American, and he
avoided abstruse philosophies of government or conduct. . . . His religion,
too, notwithstanding its Calvinist antecedents, was practical. He was religious
rather than devout. . . . The American's attitude toward culture was at once
suspicious and indulgent. Where it interfered with more important activities,
he distrusted it; where it was the recreation of his leisure hours or of his
womenfolk, he tolerated it. . . . The sense of equality permeated the Ameri-
can's life and thought. . . . The American was good natured,. generous,
hospitable, and sociable. . . . Carelessness was perhaps the most pervasive
and persistent quality in the American. . . . The American's attitude toward
authority, rules, and regulations was the despair of bureaucrats and dis-
ciplinarians. . . . The American was at once intelligent and conservative,
independent and reliable. . . . The American was romantic and sentimental.
. . . The American had a strong sense of fair play. . . . 9

All of these qualities were undoubtedly possessed by some Ameri-
cans, and some perhaps were statistically descriptive of most Americans.
But in Commager's book, they comprise a single autonomous superbeing
called "The American," a creature who appears to possess not merely
a set of normative mental characteristics, but a mind and will of its own.
Commager's use of personification makes it impossible for a reader to
distinguish a rhetorical device from a conceptual structure. Moreover, it
made it difficult for Commager himself to do so.

A second and similar form of the pathetic fallacy was recognized
more than sixty years ago by G. K. Chesterton, who called it "the fallacy
of the young nation."

It is a childish blunder, built upon a single false metaphor. I refer to the
universal modern talk about young nations and new nations; about America
being young, about New Zealand being new. The whole thing is a trick of
words. ... Of course we may use the metaphor of youth about America or
the colonies, if we use it strictly as implying recent origin. But if we use it
(as we do use it) as implying vigour or vivacity, or crudity, or inexperience,
or hope, or a long life before them, or any of the romantic attributes of
youth, then it is surely as clear as daylight that we are duped by a stale figure
of speech. 10

9. Henry Steele Commager, The American Mind, An Interpretation of American
Though! and Character Since the lS80's (New Haven, 1950), pp. 3-40, passim.

10. G. K. Chesterton, Heretics (New York, 1905), pp. 256-7. Attacks upon anthro-
pomorphism were not, of course, inaugurated by Chesterton. Other examples were
identified as early as the sixth century B.C., by Xenophanes of Colophon, who was an
outspoken critic of the anthropomorphism and anthropopathism in Greek religion.

192

EXPLANATION

Chesterton was mistaken in one respect — anthropomorphic ideas
of nations and cultures are not merely semantical tricks but serious
substantive errors. There have been many mistakes of this sort in historical
writing — both in the work of traditional historians, such as Hippolyte
Taine, and in the monographs of sophisticated social scientists, such as
Karl Deutsch, who has written about the "wills" of nations. Deutsch
defined those national "wills" as "the set of constraints acquired from
the memories and past experiences of the system, and applied to the
selection and treatment of items in its later intake, recall and decisions."
But nations do not make decisions — only people do. Sometimes people
may tend to make similar decisions within a national group, but that
kind of normative decision making is very different from Deutsch's col-
lective national will. 11

A third subspecies of the pathetic fallacy appears in many absurd
modern attempts to psychoanalyze historical events, institutions, cultures,
or nations. An English social scientist, Geoffrey Gorer, has explained the
historical relationship between Anglo-America and Europe in terms of a
national Oedipus complex, in which "England, the England of George
III and Lord North, takes the place of the despotic and tyrannical father,
the American colonists that of conspiring sons." 12 In Mr. Gorer's
interpretation, the Revolution was a great archetypical event, which has
been perpetually re-enacted in the psyches of American patriots ever
since. 13 The mother in this unhappy family is difficult to locate. Sometimes
she seems to be the Statue of Liberty, and sometimes the Republic itself.
If she is either of these things, then there would appear to be more than a
little sexual and generational confusion in Gorer's rendering of American
history, for by his logic the son becomes the father of his own mother.
Social scientists have turned up some curious family relationships in
their researches, but none as odd as this one.

Gorer is by no means alone in his errors. A learned student of Freud,
Manfred Guttmacher, has done much the same thing, on the same sub-
ject. And Max Lerner has solemnly written of a national compensation
for "the sacrificial slaying of the European father." 14

11. Karl Deutsch, Nationalism and Social Communication (New York, 1953), p. 151.
For an example of anthropomorphism in traditional historiography, see H. A. Taine,
The French Revolution, 3. vols. (Gloucester, Mass., 1962), 2:358; and see also Gordon
Shepherd, The Austrian Odyssey (London, 1957), pp. 3-22, passim, in which a
rounded personality is attributed to Austria itself, or herself or himself.

12. G. Gorer, The American People: A Study in National Character, rev. ed. (New York,
1964), pp. 29-30.

13. This, by the way, is a fine example of the fallacy of archetypes, discussed above.

14. Max Lerner, America as a Civilization, 2 vols. (New York, 1961), 1:28.

FALLACIES OF MOTIVATION

193

Gorer also applied the same method to the problems of Russian
history, explaining the success of Bolshevism in terms of the child-rearing
practices which are inflicted upon Young Russia. One critic has com-
mented that "there are more steps in any logic that leads from, say,
swaddling to the Politburo than this sort of explanation dreams of." 15

The author of that remark, Donald B. Meyer, has made very good
sense in his attempt to distinguish legitimate from illegitimate uses of
psychoanalytic analysis in history. "Efforts to transport vocabulary, to
locate a social unconscious, a collective ego, a community superego, to
equate social events with processes discovered by clinicians in individual
dynamics, are no good; these are analogies as dubious as analogies drawn
from Newton and Darwin," he has written. Psychoanalytic history must
be biographical in its orientation. It must deal with individuals. Meyer
hastens to add,

It hardly follows from this, however, that psychoanalysis, even though it
might be thoroughly adequate for the study of individuals, has no relevance
to those social entities — institutions, states, classes, styles, cultures, groups,
parties, churches, ideologies — which historians are anxious to illuminate. It
follows rather that psychoanalytic biography constitutes a perspective, or a
focus, from which history can organize all its narratives. . . . What do given
institutions, states, styles, churches, etc., mean for the selves involved in
them? 18

The distinction is wire-drawn by Meyer, but it is indispensable to any
attempt to use psychoanalytic analysis in history. If a historian wishes to
work with a group, he must remember that his method applies to the
individuals who compose the group, and not to the group itself. We might
conclude with a wise comment by Richard Pares. With nice British
understatement, he remarked that "the collective unconscious is some-
what overrated.""

The apathetic fallacy is denned by Arnold Toynbee as the error
of "treating living creatures as though they were inanimate." 18 Leaving
Toynbee's animus out of account, this form of error might be more
satisfactorily specified as treating rational men as if they were not ra-
tional. One familiar form of this fallacy is the doctrine of historical

15. Donald B. Meyer, "A Review of Young Man Luther," in Bruce Mazlish, ed.,
Psychoanalysis and History (Englewood Cliffs, N.J., 1963), p. 178.

16. Ibid., pp. 177-78, 174-80.

17. Richard Pares, "Human Nature and Politics," in The Historian's Business and
Other Essays (Oxford, 1961), p. 34.

18. Toynbee, A Study of History, 1:8.

194

EXPLANATION

materialism. Another is classical Freudian theory. The fallaciousness of
these schools of thought is succinctly explained by W. H. Walsh:

Marxists and Freudians, in their different ways, have taught us all to look
for non-rational causes for ideas and beliefs which on the surface look per-
fectly rational, and have convinced some that rational thinking as such is an
impossibility. But though we cannot (and should not) return to the naive
confidence of our grandfathers in these matters, it must none the less be
pointed out that the anti-rationalist case here cannot be stated without con-
tradiction. It undermines not only the theories of which its proponents dis-
approve, but itself as well. For it asks us to believe, as a matter of rational
conviction, that rational conviction is impossible. And this we cannot do. 18

Neither historical materialism nor psychic determinism can be
proved or disproved by appeals to historical fact. They are metahistorical
doctrines which transcend empiricism. But they cannot transcend their
own logical assumptions, which are simply suicidal, as Mr. Walsh in-
dicates. The complaint against these delusions consists not in the fact
that they are empirically false but rather in the fact that they are logically
absurd. I can see no room for compromise on this point.

The logical deficiencies of historical materialism have been discussed
elsewhere at great length. They surely require no refutation here. 20 But
there is no adequate discussion, to my knowledge, of the deficiencies of
psychic determinism in historical thought. The fallacy appears in several
works of otherwise distinguished scholarship. It is dysfunctional to the
interpretation of E. R. Dodds, in an excellent but imbalanced book on
the Greeks and the irrational. The author approvingly quotes William
James's remark that "the recesses of feeling, the dark, blinder strata of
character, are the only places in the world in which we catch real fact
in the making." 21

Some of the more picturesque examples of the same error appear
not in monographs but in metahistorical interpretations. One extreme
specimen is Gordon Rattray Taylor, Sex in History (New York, 1954).
Taylor believes that "Eros and Thanatos permeate every compartment of
human activity," 22 a proposition which is defensible. But Taylor seeks to

19. W. H. Walsh, Philosophy of History (New York, 1960), pp. 102-3.

20. An excellent critique is H. B. Acton's The Illusion of the Epoch (London, 1955).
See also M. M. Bober, Karl Marx's Interpretation of History, 2d ed. (Cambridge,
1950). The extent to which Marx was consistent in the commission of this fallacy is
of course highly controversial. See Erich Fromm, Marx's Concept of Man (New
York, 1961). But Fromm's introductory interpretation is seriously strained, as a
rounded interpretation of Marx's writings in this respect.

21. E. R. Dodds, The Greeks and the Irrational (Berkeley, 1951), p. 1; the same
assumption appears in Binion, Frau Lou.

22. V. 17.

FALLACIES OF MOTIVATION

195

demonstrate not merely "how closely attitudes to sexual matters interlock
with other social attitudes," but also how they "dictate them." All histories
which contradict this thesis are appropriately denounced as "emascu-
lated." If all thought is determined in this way, then Taylor's thought must
be determined too. And if it is, then it is difficult to imagine how his
historical inquiry is possible. In short Taylor's work is not empirically
deficient but logically self-destructive. 23

The apathetic fallacy also occurs in fragmented forms, when histori-
ans are unable to empathize sufficiently with specific human subjects.
Consider, for example, Richard Hofstadter's collection of essays, pub-
lished as The Paranoid Style in American Politics (New York, 1965).
William Buckley has protested that, in the work of Hofstadter, moderate
liberalism is analyzed, but radical conservatism is diagnosed. The
complaint, I think, is fairly made. The thought of Barry Goldwater
is no more paranoid than the thought of Hofstadter. If Goldwater's
conservatism is "fundamentalist," what shall we say of a scholar who
regards other peoples' ideologies as symptoms of a personality disorder?
Surely, there are paranoid conservatives in America, and paranoid liberals
as well. But conservatism itself, however immoderate, is not merely a form
of irrationality. To categorize conservatives in those terms is to resort to
a kind of motivational special pleading. It is also to commit the apathetic
fallacy, in a particularized form.

The idealist fallacy consists in interpretations of human con-
duct which rest upon a conception of man as Homo sapiens in a narrow
and exclusive sense. It is a double error, for it converts homines into
Homo, and Homo into Homo sapiens. The absurdity of this reduction
should be readily apparent. If any individual Homo is sapiens in some
degree, he is also furens, and prudens, diligens, and odiens, ludens, faber,
amator, and more. To isolate merely the rational component of human
existence is to falsify both humanity and rationality. Gerald W. Chapman
observed that "when history is reduced to intellectual environments . . .
expository critics risk inventing an 'intellectual man' quite as arbitrary as

23. The same mistake appears in Herbert Marcuse's Eros and Civilization (Boston,
19SS), which, with other work by the same author, seems to be an attempt to combine
the metaphysical determinism of Hegel, the economic determinism of Marx, and the
psychic determinism of Freud with a plea for human freedom! See also Norman
Brown, Life Against Death: The Psychoanalytic Meaning of History (Middletown,
Conn., 1959), which also combines the apathetic fallacy with the fallacies of meta-
history (cf. Arthur C. Danto, Analytical Philosophy of History [Cambridge, 1965],
chap. 1).

196

EXPLANATION

'economic man' in classical economics or 'political man' in classical
utilitarianism." 24

The idealist fallacy occurs both in general and particular forms.
The most prevalent general form is the so-called "neo-idealism" of R. G.
Collingwood, which is commonly understood to consist in the following
propositions:

1. All history is the history of thought . . . not only of thought but of
reflective thought, that is, one which is performed in the consciousness that it
is being performed, and is constituted what it is by that consciousness.

2. The history of thought, and therefore all history, is the re-enactment
of past thought in the historian's mind.

3. There is nothing other than historical thought itself by appeal to
which its conclusions may be verified. 25

Collingwood's conception of history is one in which thoughts alone
are events and in which an historian knows them by a process of rethink-
ing. He asserts that "the historian is not interested in the fact that men
eat and sleep and make love and thus satisfy their natural appetites." 26
But this is simply mistaken. Historians are interested in these aspects of
human behavior, and increasingly so.

And as to "rethinking," Collingwood's method calls to mind the
New England cod fisherman in Kipling's Captains Courageous:

When Disko thought of cod he thought as a cod. . . . Disko Troop stared
forward, the pipe between his teeth, with eyes that saw nothing. As his son
said, he was studying fish — pitting his knowledge and experience on the
Banks against the roving cod in his own sea. ... So Disko Troop thought of
recent weather, and gales, currents, food, supplies, and other domestic ar-
rangements, from the point of view of a twenty pound cod; was in fact, for
an hour, a cod himself, and looked remarkably like one. 27

To require a historian to rethink Brutus's thought before he killed
Caesar is to require him to become Brutus. And this he cannot do, any
more than Disko Troop could convert himself into a twenty-pound cod.

24. Gerald W. Chapman, Edmund Burke: The Practical Imagination (Cambridge,
1967), p. ix.

25. R. G. Collingwood, The Idea of History, ed. T. M. Knox (New York, 1946),
pp. 215, 308, 215, 243. Collingwood's argument, in the epilegomena to this classic
work, is a sore trial to a serious reader. The Idea of History was published posthum-
ously, from several separate and inconsistent drafts. Many attempts at explication
and systematization have recently appeared: among them, William Dray, Laws and
Explanation in History (Oxford, 1957); William Debbins, Essays in the Philosophy of
History, 2d ed. (New York, 1965); Alan Donaghan, The Later Philosophy of Colling-
wood (Oxford, 1962); and Louis O. Mink, "Collingwood's Dialectic of History,"
History and Theory 7 (1968): 3-37.

26. Collingwood, The Idea of History, p. 216.

27. Rudyard Kipling, Captains Courageous (London, 1897), pp. 47, 103-4.

FALLACIES OF MOTIVATION

197

For Brutus did not merely think different things than Collingwood
thought — he thought them differently. The whole idea is antihistorical,
antiempirical, and absurd. It is true, of course, that some history is the
history of thought and that some kinds of thoughts can be re-enacted in
the historian's mind. But some thought which interests historians cannot
be separated from feeling, or from thinking structures which exist within
limits of time and space. Collingwood's method, strictly applied, would
exclude not merely the nonintellectual problems in which historians are
actually interested but also many intellectual problems, which are char-
acteristically neither rational nor irrational, but transrational.

The epistemological issues raised by idealism are remote from the
daily thought of historians, but not from their daily work. The idealist
fallacy is a very real occupational hazard of intellectual history — a danger
from which not even the best are immune. None was better than Perry
Miller, whose studies of New England Puritanism are a landmark of
historical literature. But great though they are, these works remain deeply
flawed by the fallacy of idealism.

One problem for an idealist epistemology is the group phenomenon.
Can one rethink the thought of a collectivity? Only, it seems, by conjur-
ing up the fiction of a "corporate mind," as Collingwood called it. Perry
Miller did this explicitly in the first volume of The New England Mind,
in which he wrote, "I have taken the liberty of treating the whole litera-
ture as though it were a single intelligence, and I have appropriated il-
lustrations from whichever authors happen to express a point most
conveniently." 28

Many magnificent insights accompanied this impressionist method.
But a serious flaw was embedded in it. New England Puritanism was an
entity, but it was not an intellect. It was a cluster of many thousands of
intellects. There were important normative patterns of behavior, but also
a wide range of significant variations. Miller's method entirely prevented
a clear recognition of norms, or variations.

The second volume of The New England Mind displays another
major deficiency of the idealist method. In a preface to the paperback
edition of this work, Miller disparaged "such topics as ships, trade routes,
currency, property, agriculture, town government and military tactics."
In an argument which would have warmed the cockles of Collingwood's
heart, Perry Miller insisted that "while indeed these kinds of activity re-
quire an exercise of a faculty which in ordinary parlance may be called
intelligence, such matters are not, and cannot be made, the central theme

28. Perry Miller, The New England Mind: The Seventeenth Century, 2 vols. (Cam-
bridge, 1939-53), l:vii.

198

EXPLANATION

of a coherent narrative. They furnish forth at their worst mere tables of
statistics, on the average meaningless inventories, and at their best only
a series of monographs." 2 * This argument is not only deficient in itself,
but dysfunctional to Miller's work, for Volume Two, by the nature of its
thesis, requires a good deal of discussion of these topics which the author
despised.

Third, both volumes of the New England Mind tend, in the fashion
of idealist history, to consider narrowly rational thought. Another student
of Puritanism, Alan Simpson, has complained of Miller's work,

He has told us too much about the Puritan mind and not enough about the
Puritan's feelings. If the seventeenth century Puritan, with his formal train-
ing in scholasticism, usually tries to give a rational account of his faith, it is
the stretched passion which makes him what he is. They are people who suf-
fered and yearned and strived with an unbelievable intensity; and no super-
structure of logic ought to be allowed to mask that turmoil of feeling. 30

This turmoil of feeling cannot, ever, be re-enacted in the mind of a
historian. It can only be studied, in its behavioral expressions. And
Puritanical reasoning cannot be separated from it, without gross inter-
pretative error. Miller made much — too much — of fine-drawn dichot-
omies, but not enough of the emotional cement which was the inner bond
of Puritan belief.

A fourth common consequence of the idealist fallacy appears in
Miller's monograph on Roger Williams. In this work, the author at-
tempted, in effect, to re-enact Roger Williams's thought in his own mind.
But if a historian is to operate merely by this method, how does he
measure his success? By some sort of intuited congruence of one thought
to another? Collingwood had no answer to this problem. He was true
to the logic of his argument when he asserted that "I am now driven to
confess that there are for historical thought no fixed points thus given:
in other words, that in history, just as there are properly speaking no
authorities, so there are properly speaking no data." 31 Nothing would
bind a historian to his subject, but a gossamer web of intuited under-
standing. Miller trusted to this method sans method in his understanding
of Roger Williams and paid a price for it. His thesis is wonderfully co-

29. Perry Miller, The New England Mind, From Colony to Province (Boston, 1961),
Preface. Perry Miller began this book with an epigraph from Jeremiah: "And the
Lord put forth His hand, and touched my mouth." Vox Milleri, vox Dei est?

30. Alan Simpson, Puritanism in Old and New England (Chicago, 1955), p. 21.

31. Collingwood, The Idea of History, p. 243. An example of the consequences which
flow from this difficulty is Collingwood's explanation of King Arthur's behavior,
which he reached by attempting to rethink Arthurian thoughts. The result has been
described as "a new Arthurian legend, worthy to stand beside the inventions of Tenny-
son and Geoffrey of Monmouth," by Mortimer Wheeler in Journal of Roman Studies
29 (1939), 87-93, reviewing R. G. Collingwood and J. N. L. Myres, Roman Britain and
the English Settlements (Oxford, [1936]).

FALLACIES OF MOTIVATION

199

herent; he successfully rethinks thoughts that Williams might have
thought. Miller argued that Williams and the leaders of Massachusetts
quarreled primarily over a method of biblical exegesis. Specifically, he
believed that Williams favored a typological interpretation of the Bible,
which threatened to destroy the very foundations of Puritan thought.
But unfortunately for his thesis, there is a good deal of evidence that
typology was favorably regarded by those Puritans who banished him
from the Bay Colony. In short, by adopting a nonempirical idealist
method, Miller permitted himself to become committed to an intuited
interpretation which is empirically false. 32

Finally, historical idealism is antihistorical in its very nature, for
it would make a historian and his subject contemporaries. This actually
happens in Perry Miller's biography of Jonathan Edwards. Miller was
guilty here of the fallacy of the counterquestion, too, for in a revision
of Parrington's interpretation of Edwards, he stood Parrington on his
head. Parrington had made Edwards into a living anachronism — a con-
temporary of Cromwell. Miller overreacted and made Edwards into a
contemporary of Paul Tillich, or Marcel Proust. 33 This error was sus-
tained by his idealist method, which permitted him to move Jonathan
Edwards through time, without being aware of the mistake he was
making.

An idealist could conceivably claim, even assuming all these com-
plaints against Miller's work are well founded, that its undoubted ex-
cellence is a vindication of idealism, fallacy or not. If Perry Miller's
scholarship rests upon a fallacy, then one might wish that logical dis-
orders were contagious. I am inclined to disagree. The power of Miller's
thought is perfectly consistent with a better method — a method which
would have corrected deficiencies without destroying strengths. Every-
thing that Miller did right could have been done better in an inquiry
which took a more rounded view of human beings and a less naive view
of their condition.

A slightly different form of the idealist fallacy consists in a pre-
sumption of rationality in human behavior. One historian of ideas, C. B.
Macpherson, has elevated this assumption into an explicit method. In a
study of political thought from Hobbes to Locke he wrote, "I have found
it a fruitful hypothesis that each of the thinkers tended to be consistent,
or (which comes to the same thing) was consistent within the limits
of his vision." 34

32. Sacvan Bercovitch, "Typology in Puritan New England: The Williams-Cotton
Controversy Reassessed," American Quarterly 29 (1967): 166-91.

33. Perry Miller, Jonathan Edwards (New York, 1959), pp. 190, 289, 315, passim.

34. C. B. Macpherson, The Political Theory of Possessive Individualism: Hobbes to
Locke (Oxford, 1962), p. 7.

200

EXPLANATION

Macpherson used this method with a view to establishing with some
precision the limits of vision in his subjects. He hastened to add that it
was "no more than a useful approach." But I take it to be a very danger-
ous one. Surely many thinkers have been inconsistent within their own
limits. A presumption of logical consistency is as unjustified as a pre-
sumption of the opposite.

A final example of the idealist fallacy suggests still another varia-
tion. It occurs in the following assertion by a biographer of Ivan the
Great. J. L. I. Fennell writes,

Ivan III, more clearly than any of his predecessors or followers on the grand
princely throne of Moscow, knew precisely where he was going. He knew
his goal, the means at his disposal, the obstacles to be encountered. He never
over-estimated his own strength or under-estimated that of his enemies. His
cold reasoning told him just how far he could abuse the freedom of his sub-
jects and tamper with the sanctity of religious institutions. He never fought
a war for the sake of fighting, sought a friendship from altruism, or dis-
graced a subject through spite. All the deeds of this dedicated, hard-headed
ruler and shrewd diplomat were directed towards one goal only. 35

I do not know for a fact that this characterization is false. Every
account of Ivan III which I have seen speaks of his steely realism. Not
much seems to be known of Ivan's personality. Fennell writes, "Seldom
can a man have reigned for so long and achieved so much, and left so
little impression on his contemporaries. Almost nothing is known of his
personal qualities or of his private life." 30 Nevertheless, the cold realism
which Fennell attributes to Ivan, and the Czar's alleged awareness of
precisely where he was going, seems more than a little exaggerated. It is
an established fact that Ivan spent many hours in his life so helplessly
intoxicated that he did not know precisely where anyone was going. One
suspects that there must have been other distractions, and maybe a few
delusions as well.

The fallacy of the one-dimensional man selects one aspect of
the human condition and makes it into the measure of humanity itself.
There are many variations, of which perhaps the most prevalent is still
the fallacy of the political man. In one of its forms, this fallacy mistakes
people for political animals who are moved mainly by a desire for power.
It reduces the complex psychic condition of men merely to their political
roles and shrinks all the components of the social calculus to a simple
equation of power, ambition, and interest.

35. J. L. I. Fennell, Ivan the Great of Moscow (New York, 1962), p. 18.

36. Ibid., p. 354.

FALLACIES OF MOTIVATION

201

The classical example is provided by Leviathan and Behemoth,
both by Thomas Hobbes. In Leviathan, that excellent philosopher made
the error as explicit as it ever can be. "In the first place," he declared, "I
put for a general implication of all mankind, a perpetual and restless
desire of power after power, that ceaseth only in death." He allowed for
other restless desires, but most were also cast in political terms, and
subordinated to the primary one. 37

Hobbes tried his hand at history in Behemoth: The History of the
Causes of the Civil Wars of England, and of the Counsels and Artifices
by Which They Were Carried On from the Year 1640 to the Year 1660.
History, for Thomas Hobbes, was "a heap of evils." And historiography
was "a short narration of the follies and other faults of men." 38 When
he came to assess the cause of the English Civil Wars, all the soaring
theological aspirations of the contending parties were reduced to the
following proposition:

The seducers were of divers sorts. One sort were ministers; ministers, as they
called themselves, of Christ; and sometimes, in their sermons to the people,
God's ambassadors; pretending to have a right from God to govern everyone
his parish, and their assembly the whole nation. Secondly, there were a very
great number, though not comparable to the other, which notwithstanding
that the Pope's power in England, both temporal and ecclesiastical, had been
by Act of Parliament abolished, did still retain a belief that we ought to be
governed by the Pope, whom they pretended to be the vicar of Christ, and,
in the right of Christ, to be the governor of all Christian people. 39

The lofty return to political first principles by writers of that
troubled generation came down merely to this, in Hobbes's understand-
ing: "To these follies I might add the folly of those fine men, which out of
their reading of Tully, Seneca, or other anti-monarchies, think them-
selves sufficient politicians, and show their discontents when they are
not called to the management of the state, and turn from one side to
another upon every neglect which they fancy from the king or his
enemies." The Revolution itself, in Hobbes's pages, was but "a circular
motion of the sovereign power through two usurpers, from the late king
to this his son." 40

Hobbes's interlocutor in Behemoth observed at one point, "If this
be true, it is impossible any commonwealth in the world, whether mon-
archy, aristocracy, or democracy, should continue long without change,
or sedition tending to change, either of the government or governors."

37. Leviathan, ed. Michael Oakeshott (Oxford, 1955), pt. I, chap. 11, p. 64.

38. Behemoth, in English Works (1840), 6:270, 309.

39. Ibid., 6:167.

40. Ibid., 6:358, 418.

202

EXPLANATION

Hobbes replied darkly, "It is true; nor have any of the greatest com-
monwealths in the world been long free from sedition." 41

Thomas Hobbes, like an English mastiff, did grimly set his jaws
upon a piece of the truth about humanity, but it was a piece from the
hind end. He made men into beasts, braying and biting and trying their
titles with their teeth. His political man was scarcely a man at all, but
rather an untamed animal.

No historian of the twentieth century has, to my knowledge, built
such an idea into such a system. But many historical interpretations are
sometimes informed by a comparable assumption about a particular
political man. It is often a variable thing, which appears and disappears
in a single work. There are many examples in the historiography of the
early republic. Scholars of a Federalist sympathy have committed the
fallacy of the political man in their characterizations of Jeff ersonians. And
Jeffersonian historians, in turn, have introduced it to explain the behavior
of Alexander Hamilton. An example of the latter is the writing of
Adrienne Koch, who, in a recently published collection of essays called
Power, Morals and the Founding Fathers, included a chapter called
"Hamilton and the Pursuit of Power." The first Secretary of the Treasury
becomes a power-worshiping fanatic, whose career is explained by this
"all-consuming passion." His end is seen as a kind of poetic justice: "It
was as if fate had come to punish him for his grave defect of character
and consequent crimes with an early and violent death," she wrote
piously. 42

Adrienne Koch's monistic interpretation of a particular single indi-
vidual in terms of "a perpetual and restless desire for power after power,
that ceaseth only in death," is every bit as reductive as Hobbes's similar
opinion of mankind in general. Many students of Hamilton's career —
enemies as well as friends, and objective observers besides — would surely
agree that he was consumed by many other passions besides a passion
for power. Mrs. Reynolds could have named one. Moreover, Hamilton
was also consumed by many altruisms and ideas and interests, some of
which had nothing whatever to do with power in any meaningful sense
of the word.

A second general species of the fallacy of the political man is that
which reduces men merely to citizens, or subjects, or voters, or party
functionaries, or officeholders. There was a good deal of this in many
monographs published by constitutional and political historians fifty
years ago, as well as in many general interpretations of American his-

FALLACIES OF MOTIVATION

203

tory. It still appears more often than it should. Carl Prince's book New
Jersey's Jeffersonian Republicans (Chapel Hill, 1967) is a case in point.
The leading figures in this volume appear as political men on the make,
mindless automatons programed for a maximization of political power.
I regret that I did something of the same thing in a study of the Federal-
ists. The complaint here is not about a special interest in political
history but rather about the way in which that interest is developed, with-
out sufficient attention to its cross-connections with other aspects of
human conduct.

Other examples are discussed in an excellent analysis of motivation
in writing on the Reconstruction era. What were the motives of Radical
Republican politicians for enacting Negro suffrage? The answer, in many
interpretations, is flat, simple, and monistic. Republican politicians
wanted the political support of the Negro voters! 43 This profundity can
be taken not merely as a truth, but as a truism. And yet, it is all that a
good many historians wish to know.

There are many other forms of the fallacy of the one-dimensional
man. Everybody has some acquaintance with the fallacy of economic
man — an idea steadfastly asserted by shabby scholars with holes in their
socks. And besides Homo economicus, there is Homo religiosus and
Homo sociologicus, and many more. 44 Some historians have even dis-
covered Homo historicus. Bruce Mazlish appears to believe that the
meaning of history consists in man's developing historical consciousness.
What next! 45

The fallacy of the universal man falsely assumes that people
are intellectually and psychologically the same in all times, places, and
circumstances. It is an error which has ruined the designs of innumerable
Utopians, revolutionaries, schematizers, prophets, preachers, psychiatrists,
mystics, cranks, and social scientists of very shape and hue. Every
unitary solution, without exception, which has ever been proposed as a
panacea for the hopes and misfortunes of mankind, has been fatally
flawed by this fundamental fallacy.

People, in various places and times, have not merely thought differ-
ent things. They have thought them differently. It is probable that their

43. Lawanda and John H. Cox, "Negro Suffrage and Republican Politics: The Problem
of Motivation in Reconstruction Historiography," Journal of Southern History 33
(1967): 303-30, discuss this error.

44. Ralf Dahrendorf, Homo Sociologicus: Ein Versuch zur Geschichte Bedeutung
und Kritik der Kategorie der sozialen Rolle (Cologne, 1960).

45. Bruce Mazlish, The Riddle of History (New York, 1966).

204

EXPLANATION

most fundamental cerebral processes have changed through time. Their
deepest emotional drives and desires may themselves have been trans-
formed. Significant elements of continuity cannot be understood without
a sense of the discontinuities, too.

The problem is immensely difficult, and highly doubtful. There is no
primary evidence for the past thought of any man, at any time in his-
tory — though psychologists and biologists are beginning to discover paths
into this inner wilderness. But there is accumulating evidence of expres-
sions of thought and feeling which make no sense unless we allow a wide
latitude for change in the nature of cerebral activity through space and
time. The range of this change is as obscure as its nature. But its exist-
ence is, in my opinion, a historical fact which is established beyond a
reasonable doubt. 48

The fallacy of the universal man appeared in much eighteenth-
century historiography. David Hume declared that "Mankind are so
much the same, in all times and places, that history informs us of nothing
new or strange in this particular. Its chief use is only to discover the
constant and universal principles of human nature." 47 Voltaire was
equally of the opinion that "Man in general has always been what he is
now; this does not mean that he has always had fine cities . . . and
convents full of nuns. But he has always had the same instinct which
leads him to find satisfaction in himself, in the companion of his
pleasures, in his children, in his grand children and in the work of his
hands. Here is something that will never change from one end of the
world to another." 48 But "man-in-general" changed from one wing of the
Enlightenment to another. Much of what Voltaire described as univer-
sally present in human nature was, for example, conspicuously absent
from the nature of Jean Jacques Rousseau, who derived mostly pain
from the companions of his pleasure and gave his own children to an
orphanage. Rousseau and Voltaire themselves are two clear examples
of men who did not merely think different things but thought them
differently.

In our own time, there are many other examples of the fallacy
of the universal man. Its progress has been much encouraged by
two hopeful tendencies in the modern world. The first is a powerful
reaction against the fatal fallacy of racism. The repudiation of this
bloody error by most historians, and many others, is surely cause for
rejoicing. But some have overreacted and insufficiently allowed for the

46. For a preliminary attempt to deal with this problem in some detail see J. H. Van
den Berg, The Changing Nature of Man (New York, 1961).

47. Hume, Essays (1767), II, 94.

48. Voltaire, Selected Writings, ed. J. H. Brumfitt (New York, 1963), pp. 260-62.

FALLACIES OF MOTIVATION

205

existence of cultural differences among men. Consider, for example, the
historiography of Negro slavery in America. Historians have disputed
many aspects of that mournful institution. Most bitter, and most funda-
mental, have been their quarrels over the nature of men who were
slaves. In 1918, a very great historian of this problem operated upon an
explicitly racist premise. "The slaves were negroes," wrote Ulrich B.
Phillips, "who for the most part were by racial quality submissive rather
than defiant, light-hearted instead of gloomy, amiable and ingratiating
instead of sullen." 49

More recently, the author of another standard history of the same
subject began with a very different assumption, which was actively anti-
racist. "I have assumed," wrote Kenneth Stampp, "that the slaves were
merely ordinary human beings, that innately Negroes are, after all, only
white men with black skins, nothing more, nothing less." 50

Therein lies our fallacy. Stampp assumes, throughout his book, that
an African Negro slave responded to his predicament much as Stampp
himself might have done. It is as if that white liberal professor of history
who worked in twentieth-century Berkeley, California, were somehow
shackled, and put upon the block, and sold to a nineteenth-century
plantation owner in Bibb County, Alabama (a thought which, by the
way, might not seem so ridiculous to the incumbent governors of those
two states). In the pages of Stampp's Peculiar Institution, the mark of
Cain has disappeared from the brows of his protagonists. But every
Negro is, as it were, Stampped into a white liberal stereotype which may
be equally false.

It is more sensible, in my opinion, to operate upon a different
assumption, which was more closely approximated in a brilliant essay
by Stanley Elkins, published since Stampp's book appeared. Though
there are surely many flaws in Elkins's approach, some of which are dis-
cussed in this book, he has opened most impressively one important line
of inquiry in his suggestion that the nature of the acculturation process
through which Negro slaves passed from one culture into another, and
the role expectations which were ruthlessly enforced upon them, may
have served to create a pattern of psychological and intellectual behavior
which was very different from that of whites in this country, a pattern
which was so deeply and powerfully ingrained in the nature of American
Negroes that they did not begin to break out of it in significant numbers
until a century after emancipation. Maybe that behavior pattern re-
sembled in a superficial way the qualities which Ulrich Phillips and many

49. Ulrich B. Phillips, American Negro Slavery (Baton Rouge, La., 1966), pp. 341-42.

50. Kenneth Stampp, The Peculiar Institution (New York, 1965), p. vii. This work
was first published in 1956.

206

EXPLANATION

others described in Negroes whom they knew — docility, dependence,
irresponsibility, imitativeness. Beneath these mannerisms, there may
have been a more fundamental set of psychic determinants and dy-
namics. Perhaps this behavior pattern, though not racially inherent in the
nature of black men, nevertheless inhered to his cultural condition
throughout his period of bondage, and for many years thereafter. Many
Negroes today have clearly broken out of this psychic pattern; the black
militant movement is a striking case in point. But maybe many others
are still caught up in it and must still be emancipated from the worst
shackles of all — shackles of the mind.

This view of Negro history is understandably unpopular among
militant blacks. They insist that such personality stereotypes as Ulrich
Phillips described and Stanley Elkins has begun to explain were merely
social masks, if they existed at all — masks that blacks wore in the
presence of whites. But maybe these manners ran deep in the minds of
the enslaved.

Elkins seems to suggest in his book that the institution of slavery in
the United States was such that it might have made Sambos of any
slave, white or black or red, African or European or American. But
maybe there were African cultural determinants which made blacks
particularly vulnerable. There is good evidence that American Indians
behaved very differently in slavery — so differently that they could not
be successfully enslaved on a large scale, though many attempts were
made. 51

To represent Negroes merely as "white men with black skins," in
Stampp's unfortunate phrase, is to deny their history. It is moreover to
disguise the full dimensions of their modern dilemma, which cannot be
dealt with merely by a neutral kind of equality of opportunity in an open
and competitive society. If it is true that Negroes, by reason of their
African inheritance, their acculturative experience, and their centuries
of bondage, do not merely think different things but think them differ-
ently, then that fact has a most profound relationship to modern policies
which have been designed to deal with the so-called race problem. The
fact has two implications, which do not easily coexist. First, any simple
attempt to assimilate Negroes into white society, on the ground that they
are really white inside, is a threat to black identity more serious even
than that which was posed by Ulrich Phillips's racism. There is no greater
disservice to a black American than to call him a white man with a black
skin. On the other hand, it may be true that Africans in America histori-
cally became people of such a nature that many of them are still un-

51. Cf. Stanley Elkins, Slavery: A Problem in American Institutional and Intellectual
Life, 2d ed. (New York, 1963), pp. 81-139.

FALLACIES OF MOTIVATION

207

equipped to function in the modem world on a basis of equality. What
would be required is some radical form of compensatory opportunity
which would provide assistance without assimilation, a helping hand
without homogenization. Stampp's well-intended liberal interpretation
of the peculiar institution is, in short, doubly objectionable. It fails to
perceive the greatest evil of slavery, which was not the use it made of
men but the kind of men it made. And it obscures the nature of the
problem with which America is presently faced, as well as offering a
false solution. It is, I believe, America's glory that no other nation, any-
where on the face of the planet, has ever attempted to create a free,
open, and equal society from such disparate materials. If the attempt
is to succeed, we must have a sufficient sense of those disparities, as well
as a respect for the integrity of the various individuals who are
involved. In such a situation, the fallacy of the universal man is not
merely a delusion but a danger. It is not, as its well-meaning proponents
imagined, a tool of international peace and brotherhood, but the very
opposite — a formidable threat to these lofty ideas.

The fallacy of the mass man, in the words of E. H. Carr, con-
fuses anonymity with impersonality. 82 It commonly happens when mil-
lions of individual living men are changed into "masses" and are thereby
deprived of individuality and life itself. John Stuart Mill wisely observed
in bis System of Logic that "Men are not, when brought together, con-
verted into another kind of substance." And yet they often are, when
brought together for the purpose of historical analysis.

A few historians have been protesting against this practice for years.
Carlyle, in his French Revolution, wrote,

With the working people, again, it is not so well. Unlucky! For there are
twenty to twenty-five millions of them. Whom, however, we lump together
into a kind of dim compendious unity, monstrous but dim, far off, as the
canaille; or, more humanely, as the "masses." Masses indeed, and yet, singu-
lar to say, if with an effort of imagination, thou follow them, over broad
France, into their clay hovels, into their garrets and hutches, the masses con-
sist all of units. Every unit of whom has his own heart and sorrows; stands
covered there with his own skin, and if you prick him, he will bleed. 53

So it was and is in history, but rarely in historiography. There is
less humanity in billions of bloodless historiographical John Does and

52. Carr, What Is History?, p. 61.

53. Thomas Carlyle, The French Revolution, Modern Library ed. (New York, n.d.)
p. 28.

208

EXPLANATION

Juan de la Cruzes than in a single prince or priest. Political historians
of a conservative persuasion have received a good deal of deserved abuse
for this detestable practice. But the error is not theirs alone. It commonly
occurs in radical social history, too. Consider a recent work by Jesse
Lemisch, on merchant seamen in colonial America. 54 Lemisch belongs
to a small sect of scholarly dissenters who are commonly called New
Left historians. He has actively repudiated the elitism of most academic
historiography, but only to stand it on its head. In the essay at hand, he
has homogenized a good many merchant seamen into a mass man who
is explicitly called Jack Tar. This abstraction is not quite as dim or
compendious as the mass men of conservative historians, but it is equally
artificial. Lemisch's Jack Tar has about as much empirical validity as
Jack and the Beanstalk — whom we might expect to meet when he gets to
agricultural workers of the colonies. His Jack Tar is quite a jolly fellow,
curiously comparable to the eighteenth-century country-gentleman ab-
stractions celebrated by conservative scholars — a little on the rough
side, but fair and firm and virtuous and manly and quick to resent
restraints upon his liberties. In all of this, the range of individual acts
and thoughts and feelings tends to disappear.

No denigration of Lemisch's work is intended, for he has carried
our understanding of an important problem far beyond all earlier
efforts. But in his writings the myth of the mass man remains — an affec-
tionate abstraction which, whatever its connotation, remains an empty
stereotyDe. A better strategy than the one Lemisch adopted — from an
empirical point of view — would surely have been to conceptualize his
problem in terms of ranges of normative behavior of individual merchant
seamen.

The fallacy of the man-mass conversely converts a singular
individual, with all his quirks and idiosyncrasies, into a collectivity of
individuals. It appears in an unfortunate stylistic habit of Louis Hartz,
who likes to couple a definite noun with an indefinite article. In his
Liberal Tradition in America, "the joy of a Dewey meets the anguish of a
Fenimore Cooper," and a good many metaphorical sparks fly, at least
in the mind of this reader. He also writes of "a Jefferson" and "a Frank-
lin," of "Harringtons and Machiavellis and Rousseaus," or "a Duke of
Wellington" and "an Abbot Lawrence," of "a Flocun" and "a Ledru
Rollin," "an Edmund Burke" and "a Thomas Paine," of "Benthams,

54. Jesse Lemisch, "Jack Tar in the Streets," The William and Mary Quarterly, 3d
ser. 25 (1968): 371-407.

FALLACIES OF MOTIVATION

209

Mills and Carlyles," of "Disraelis and Bonalds," and even of "a Calvin"
and "a Lenin." 65

A contextual examination of these monstrosities suggests that
Hartz's confusion is more than merely grammatical. It consists in the
way in which he conducted both his conceptualization and his research.
Hartz repeatedly took an available individual and enlarged him into
a group. The result is that all the Federalists become "an Ames" or "a
Hamilton," which is humbug. Having homogenized his subject in this
way, Hartz proceeds to proclaim to all the world that his subject was
homogeneous: America, it seems, was really just a Locke, to which a
Beard had lost the key.

To complain of this practice is not to protest against all general-
ization but merely against a species of false generalization in which an
individual is puffed up like a balloon and mistaken for a class of indi-
viduals. This error is common among both social scientists and historians.
It appears in the writing of behaviorists like David Easton, who has
referred to "a Walter Bagehot"; and also in the work of a traditional
historian such as F. C. Lane, who speaks of "a John Adams." 58

The historian? fallacy is suggested by William James's "psy-
chologists' fallacy," which he defines as the error of assuming that a
man who has a given psychic experience knows it, when he has it, to be
all that an observing psychologist would know it to be. 57 In precisely
the same way, the historian's fallacy is the error of assuming that a man
who has a given historical experience knows it, when he has it, to be
all that a historian would know it to be, with the advantage of historical
perspective.

One common form of this fallacy consists in the tendency of his-
torians, with their retrospective advantages, to forget that their subjects
did not know what was coming next. This sort of error is ludicrous, in
an abstract way. Imagine a letter written in France, on May 24, 1337,
which announced "the Hundred Years' War began here today." But
the historians' fallacy is often an insidious form of error, which is ex-
ceedingly difficult to recognize or correct. A case in point is the continu-
ing controversy over the surprise which we received from the Japanese

55. Louis Hartz, The Liberal Tradition in America, pp. 39, 46, 59, 107, 116, 120-21,
141, 148, and 157. Italics added.

56. David Easton, A Framework for Political Analysis (Englewood Cliffs, N. J., 1965),
p. 81; and Frederic C. Lane, "At the Roots of Republicanism," The American Historical
Review 71 (1966): 420.

57. lames, Principles of Psychology, I, 196.

210

EXPLANATION

at Pearl Harbor, on December 7, 1941. Before the bombs began to
fall, there were many clear signals of an impending attack — so many,
and in retrospect so clear, that many people have suspected some sort
of foul play. Almost every historian of this event has recorded the ac-
curate information which Ambassador Grew collected in Japan months
before Pearl Harbor Day. Everybody remembers the wealth of informa-
tion collected from magic, and the army radar operator who actually
tracked the incoming Japanese planes. But our memory does not extend
with equal clarity to many other signs and signals which pointed un-
equivocally in the opposite direction. An excellent recent monograph has
contributed considerably to our understanding of the problem by analyz-
ing it precisely in these terms. Its author, Roberta Wohlstetter, sensibly
observes that the

signals announcing the Pearl Harbor attack were always accompanied by
competing or contradictory signals, by all sorts of information useless for
anticipating this particular disaster. ... To understand the fact of surprise
it is necessary to examine the characteristics of the noise as well as the
signals that after the event are seen to herald the attack. ... In short, we
failed to anticipate Pearl Harbor not for the want of the relevant materials,
but because of the plethora of irrelevant ones. 58

Mrs. Wohlstetter's method might be studied with profit by many profes-
sional historians — and intelligence analysts. It is a very useful corrective
to a very common form of error.

The historians' fallacy appears full-blown in Louis Hartz's The
Liberal Tradition in America, which denounces Hamilton and Adams,
and Federalists and Whigs, for failing to understand their society. "They
deserve all of the criticism they have received," he writes, "but not for
the reason they have received it. Their crime was not villainy but stupid-
ity. . . . What is remarkable is how long the American Whigs managed to
endure the strange abuse of a liberal community without waking up to
the logic behind it." 89 Now, Hamilton and Adams can be accused of a
good many things, but not stupidity. If Hartz's charge says anything
coherently, then I think it consists in an indictment of the founding
fathers for failing to read the future. In retrospect, we can find clear
signs in the early republic which point to the direction of change in
nineteenth-century America. But there were also many other signs, which
pointed in different directions. When Hartz denounces the Federalists as
a pack of dunces for failing to read the right signs, he reflects not so
much upon the limits of their perspicacity as upon the limits of his own.

58. Roberta Wohlstetter, Pearl Harbor: Warning and Decision (Stanford, 1962), pp.
3, 387, passim.

59. P. 101

FALLACIES OF MOTIVATION

211

On the other hand, there are many historians who have also com-
mitted the historians' fallacy by celebrating the cerebration of the Found-
ers. Marvin Meyers, in a paper presented to the American Historical
Association in December, 1 967, came close to interpreting John Adams
in such a way that Adams recognized himself to be a Founding Father
with the clarity that we possess in retrospect. Mr. Meyers suggests that
Adams had a clear and accurate sense of the way in which the Republic
would develop after its founding, and that both of these insights were
intertwined in his motivational patterns and ideological responses. It is
as if the American revolutionary slogan were "Liberty, Equality, Pater-
nity" and the Fathers of the Republic were driven by a sense of father-
hood and by an accurate premonition of the career of their republican
child.

Still other historians' fallacies have been committed in the historiog-
raphy of the early Republic, when scholars of many different political
persuasions have analyzed the great party conflicts of the 1790s as if
the participants knew what was going on with great clarity and precision.
But the reality, I suspect, was a little like that which Herbert Feis de-
scribes in a meeting between Franklin Roosevelt and Herbert Hoover on
January 20, 1933. Feis has written,

Of the many confused scuffles it has been my professional pleasure to study,
the one that ensued is the hardest to relate with confident accuracy. It was
reminiscent of a naval engagement on a foggy night between two opposed
fleets, each ship firing a gun whenever a flash was seen, being quite as likely
to blow up a friend as an enemy. In this instance as well, the proponents
were shooting at shadows and hitting the air. 60

Feis's analysis is precisely applicable to the party battles of the
early Republic, and to many other conflicts as well — political, intellec-
tual, constitutional, military, economic, etc., etc. To apply this insight is,
however, extraordinarily difficult. Nothing is more elusive than the in-
formation which somebody did not have, and yet nothing is more
useful to a coherent and accurate historical interpretation.

Another specimen of the historians' fallacy is Philip C. Ritterbush's
Overtures to Biology: The Speculations of Eighteenth Century Natural-
ists (New Haven, 1964), a useful and important book. But a reviewer
has complained that "at least twice the author falls into one of the worst
blunders that a historian of science can commit: hindsighted contempt
for the ignorance of past investigators. Vaillant 'should not have ex-
pected' something to happen that we know (but he did not) could never
happen. Linnaeus was guilty of an 'outrageous' assumption of something

60. Herbert Feis, 1933: Characters in Crisis (Boston, 1966), p. 69.

212

EXPLANATION

that we know to be impossible." 61 The critic, Donald Fleming, needn't
have been quite so self-righteous about it. But the critical point is soundly
taken.

Still another example is Jack Lindsay, Leisure and Pleasure in
Roman Egypt (New York, 1966), which a reviewer, Ramsay Mac-
Mullen, criticizes as follows:

Though factual errors are few, and minor, yet a broader fault lies in the re-
current tendency to attribute to people of the time an informed awareness
of their own culture now matched only by some curator of Egyptian an-
tiquities. How many of us, however, know why we greet a sneeze with
"Bless you!" or why we deck a pine tree at Christmas? The man in the street
of Roman Egypt cannot have responded much more sensitively to the fasci-
nating cultural lights, meanings and connections that the author detects. 62

Curators of Egyptian antiquities, of course, are something less than
omniscient within their chosen field. There is surely much that "the
man in the street of Roman Egypt" knew about his own culture that no
scholar will ever succeed in reconstructing. But there are other analytical
advantages which operate on the opposite side, and they are a mighty
snare and a delusion, in any retrospective inquiry.

A final example of the historians' fallacy is a borderline case. An
excellent narrative historian, George Dangerfield, has written of Andrew
Jackson that "the General seemed to regard himself as an extragovern-
mental force, a special spirit, unaccountable to anyone or anything but
the nation and the frontier." 03 The meaning of this statement is a little
obscure, but it comes close to suggesting that Andrew Jackson carried
the collected works of Frederick Jackson Turner in his saddlebag.

Many historians have been conscious of the historians' fallacy. A
few, indeed, have become so obsessively conscious of it that they have
fallen into a counterfallacy. Douglas Southall Freeman made a deter-
mined effort to avoid it in his great biography of Robert E. Lee by the
desperate expedient of telling his readers no more about a given situa-
tion than Lee himself knew. According to this method, which is called
the "fog-of-war" technique by military historians, the object is to help
the reader to find a rapport with Lee by providing a common bond of
collective ignorance. It is, paradoxically, an effort to refine the clarity
of our understanding of Lee by experiencing his own confusion. But
this method, I think, is a mistake. There are certain items of information
which every reader of R. E. Lee will bring to the book, such as who won

61. The American Historical Review 71 (1965): 113.

62. Ibid., 72 (1966): 139.

63. George Dangerfield, The Awakening of American Nationalism (New York, 1965),
p. 47.

FALLACIES OF MOTIVATION

213

the Civil War and what happened at Gettysburg. To try to put the reader
in Lee's place, and to seek a sort of historiographical high fidelity by
many ingenious methodological tweeters and woofers, simply won't
work; there is an irretrievable artificiality which cannot be overcome.
Moreover, we possess many advantages in retrospection that ought not
be relinquished.

$»» If the reader is willing to accept the fallaciousness of these
putative fallacies, then perhaps he will also accept a set of affirmative
propositions which they suggest. These propositions are derived, in part,
from a reading of some recent work in motivational psychology. That
discipline has been exceedingly dynamic in past decades, as psychological
thought has developed from the "instinct" orientation of James and
Thorndike and MacDougall, through a stage in which "drives" were con-
ceptually central, to a third stage in which the predominant idea appears
to be "goal-directed" activity — which is to say, motivation in a more pre-
cise and literal sense. 84

The work of Abraham Maslow and David McClelland might be
particularly pertinent to historians' purposes and particularly valuable as
a basis for rapport between psychology and history. Maslow and McClel-
land, if I understand them, have created a new synthesis of psychological
thought, in which consciousness plays an ever larger role, though the
importance of the unconscious is not repudiated. This general tendency
is apparent in the general transformation of "instincts" to "drives" to
"motives." As such, it brings psychological theory closer to the prob-
lems historians actually face in their work. A work such as David
McClelland's The Achieving Society speaks directly to classic histori-
ographical issues and suggests the possibility of a bridge between
historical research and psychological theory. If these opportunities for
a genuine interdisciplinary interaction can be exploited, then we may
anticipate the refinement of motivational theory and motivational ex-
planation by a union of psychology of history, which might begin with
the following assumptions, all of which are the converse of the fallacies
discussed in this chapter.

1. Motives are properly understood in terms of anticipatory goal
states which are often physiologically associated in ways which we are
only beginning to understand but which are never mere reflexes of
biological or external stimuli.

64. See generally, Chalmers L. Stacey and Manfred F. Demartino, eds., Human
Motivation, rev. ed. (Cleveland, 1963), esp. pp. 5-11; and David C. McClelland, ed.,
Studies in Motivation (New York, 1955), esp. pp. 226-34.

214

EXPLANATION

2. Only intelligent individual beings have motives — not sub-intelli-
gent beings, not things, not groups. There are, however, normative mo-
tivational patterns which some individuals share with others in some
degree.

3. Every motive is learned. David McClelland writes that "it must
involve two points on an affective continuum: a present state (either
positive, negative or neutral) which redintegrates through past learning a
second state."

4. Motives have been learned differently in different times and
places, so as to require conceptualization in developmental terms and in
terms which respect the variability and mutability of particular cultural
and physical environments. Psychological normality and abnormality are
themselves temporally and culturally relative ideas which must be located
in time and space.

5. Motives are usually pluralistic in both their number and their
nature. Abraham Maslow writes, "Typically an act has more than one
motive." To this, one might add that it has motives of more than one
kind.

6. In any given motivational pattern, conscious and unconscious
motives tend to coexist and interact. Neither consciousness nor the un-
conscious should be understood as prior states. Men are neither per-
fectly rational nor perfectly irrational but imperfectly both.

7. Conscious motives and unconscious motives are themselves
pluralistic. In the realm of consciousness, a man who does something
does it for every reason he can think of, and a few unthinkable reasons
as well. Similarly to search for the motivational key in the unconscious
is to commit the reductive fallacy.

8. Motivational pluralism must not, however, become an indis-
criminate pluralism. There are what Maslow calls "hierarchies of pre-
potency" in motivation. There are many different levels of priority, in-
tensity, and specificity. Moreover, in Maslow's words, "the appearance
of one need usually rests on the prior satisfaction of another, more
preponderant need."

9. Motive sets are never, ever, inalterably fixed in a living individual
— not in the first six years, not in adolesence, and not in early adulthood.
Karl Jaspers has written that "Man ... is not what he is simply once
and for all, but is a process." 65 His psychic process may, perhaps, be
arrested for analytical purposes, but only in a fictional way. In fact, it is
always in motion.

10. Similar sets of motives express themselves in similiar ways;

65. Karl Jaspers, Man in the Modern Age (Garden City, N.Y., 1957), p. 159.

FALLACIES OF MOTIVATION

215

different sets, in different ways. But single motives can be combined in
different sets.

11. Motive sets affect, and are affected by, an entire man and
not some segment of his self, or some fragment of his social existence.
If any part is singled out for special study, it always remains a part — not
the whole.

12. An empirical understanding of normal motivation must be
derived from the study of normal men. The work of psychol-
ogists is diminished by its dependency upon the behavior of animals and
undergraduates. The work of psychiatrists is narrowly based upon
neurotic behavior, from which norms have been inaccurately extra-
polated.

13. There can be no primary, direct evidence of any past motive.
But there is a tacit logic of inference which can attain a high degree of
probable accuracy. It is a logic which in its very nature appears to
commit the fallacy of the consequent (in the form "if X, then probably
Y; Y, therefore probably X"). But this form of reasoning is a useful
tool of empirical inquiry.

Bland as these statements may seem, they contradict many — per-
haps most — historical interpretations. They cannot coexist with Marxian
historiography, or with classical Freudian theory, or with much con-
ventional wisdom of academic history. Moreover, if they are correct,
then interdisciplinary efforts in history and psychology would not take
the form of mere consumption of psychological insights by historians,
but rather a flow of fact and theory in each direction, with mutual
benefits accruing to both disciplines.

CHAPTER VIII

FALLACIES OF
COMPOSITION

It i s one of the basic characteristics of history that the his-
torian is concerned with human beings but that he does not
deal with them primarily as individuals, as does the psy-
chologist or the biographer or the novelist. Instead he deals
with them in groups — in religious groups, in cultural groups,
in ideological groups, in interest groups, in occupational
groups, or in social groups.

—David M. Potter

The subject matter of history is always men in the midst of other men —
men in collectivities and groups. A human group is something more than
merely a heap of people and something other than an organism or a
machine or a great person or an idea. A group is not exactly born, and
it does not precisely die. It has a beginning and an end, but no life cycle,
no organic pattern of growth and decay. It has no roots or branches; no
fruits or flowers; no mind or heart or soul; no cogs or gears or wheels
or levers. It does not possess a will or a personality.

Human groups exist in their own right and must be understood in
their own terms. We might define them in terms of five properties: a
finite membership of particular individuals, a regular structure of inter-
action, a normative pattern of behavior to which its members conform in
some degree, a set of functions which it performs for its constituents or
for other groups, and a sequence of development through time. An in-
numerable variety of group types have actually engaged the interest of
historians. The following are merely a few in common use.

1 . Cultural groups, subcultures, communities, civilizations, Gemein-
schaften.

2. Social groups, societies, Gesellschaften.

3. Political groups, nationality groups, states, nations, polities,

FALLACIES OF COMPOSITION

217

political cultures, republics, kingdoms, parties, factions, parliaments, con-
gresses, caucuses, conventions, cabals, camarillas.

4. Economic groups, exchange groups, investment groups, con-
sumer groups, consortiums, combinations, communes, collectives, man-
ors, plantations, farms.

5. Religious groups, religions, churches, denominations, sects, com-
munions, congregations, cults.

6. Enculturative groups, educational groups, language groups,
schools, colleges, faculties, student-groups.

7. Ideological groups, belief units, value units.

8. Reference groups, classes, castes, cliques, coteries.

9. Kinship groups, families, clans, tribes, gens, septs, connections.

1 0. Generational groups, age groups, peer groups, gangs.

1 1. Residential groups, hamlets, villages, towns, cities, suburbs.

12. Vocational groups, occupations, professions, guilds, crafts.

13. Military groups, armies, officer castes.

14. Voluntary associations, fraternities, fellowships, bands, clubs.

15. Temporary group aggregates, crowds, caravans, committees,

etc. 1

Each of these group terms has a broad range of meanings in com-
mon usage. Two anthropologists have located more than 160 different
working definitions of "culture" alone in the literature of social science. 2
Sometimes these terms refer not to a group itself but to a set of group
possessions. But always the existence of some sort of group, in the
fivefold definition given above, is at least tacitly implied.

The purposes of group study have been as various as the groups
themselves. Sometimes groups are important to historians merely as con-
texts in which individuals acted or in which particular events happened.
Other historians are increasingly engaged in the history of groups them-
selves. In political history, for example, a predominant concern with
individual acts and particular events is yielding to an interest in what are
called political cultures. These entities are set in motion and themselves
become the central object of historical inquiry. The same trend is ap-
parent in social history, which too often in the past meant merely an
interest in all the discrete happenings which political historians had

1. This list is provided for purposes of illustration, not formal classification. Though
there is a large sociological literature on the subject, no definitive taxonomy of groups
is possible, for new groups are always in the process of formation, and new group
types, too. Cf. Robert K. Merton, Social Theory and Social Structure, rev. ed. (Glencoe,
111., 1957), pp. 309-10n.

2. A. L. Kroeber and Clyde Kluckhohn, Culture: A Critical Review of Concepts and
Definitions (Cambridge, Mass., 1952).

218

EXPLANATION

ignored. Today, however, social history is becoming a sophisticated study
of the lineaments of society itself, as they have changed through time.
The new economic historians, the new historians of education, the demo-
graphic historians, the new diplomatic historians, and the new historians
of science are all moving on parallel lines.''

3. Cf. R. W. Fogel, "The New Economic History: Its Findings and Methods,"
Economic History Review, 2d ser. 29 (1966): 624-56; Allan G. Bogue, "United
States: The 'New' Political History," Journal of Contemporary History 3 (1958):
5-27; Philip J. Greven, Jr., "Historical Demography and Colonial America," William
and Mary Quarterly, 3d ser. 23 (1966): 627-34; Wilson Smith, "The New Historian
of American Education," Harvard Educational Review 31 (1961): 136—43; Kuhn, The
Structure of Scientific Revolutions, passim; Lawrence Stone, The Crisis of the Aris-
tocracy, passim; T. Scott Miyakawa, Protestants and Pioneers, passim. This new work
appears not merely in manifestoes but in monographs. It is sustained not by mere
aspiration but by solid achievement. And it is happening in every historical field. One
finds, for example, studies of social mobility in ancient China and modern America;
and monographs on authority and alienation in ancient Rome and modern Australia.
See, e.g., Cho-yun Hsu, Ancient China in Transition (Stanford, 1965); Stephan Thern-
strom, Poverty and Progress (Cambridge, 1964); Ramsay MacMullen, Enemies of the
Roman Order (Cambridge, 1966); and Michael Roe, Quest for Authority in Eastern
Australia (Parkville, Australia, 1965).

This new history is an international phenomenon, which first became quantitatively
significant in France and has since spread to other nations. One finds it in the great
French journal, Annates, and in the major work of Marc Bloch and Fernand Braudel.
Many new historical journals, mostly founded in the 1950s and 1960s, are similarly
oriented. One thinks of Past and Present (1952, Oxford), Saeculum (1950, Munich),
the International Review of Social History (1956, Amsterdam), History and Theory
(1960, Middletown, Conn.), Comparative Studies in Society and History (1958, Ann
Arbor, Mich.), and The Journal of Contemporary History (London, 1966). I am told
that the same trends are apparent in Russian journals, too — Novaia i Noveishaia
Istoriia (Modern and Contemporary History) (1958, Moscow); and Vestnik Istorii
Mirovoi (Journal of the History of World Culture) (Moscow, 1957).

The new history seems to be moving very fast in France, the United States,
and also the developing nations. In the latter, there is a curious kind of combined
development, in which exceedingly primitive attitudes toward history exist side by side
with the most sophisticated projects. In Indonesia, there are still interdisciplinary
disputes between history and magic, at the same time that there are some remarkable
projects in social and cultural history. See Soedjatmoko, ed., An Introduction to
Indonesian Historiography (Ithaca, 1965). Some groups of historians are more
laggard than others; German academic historiography may be the most backward
in the world. But even in Germany there are stirrings by both history-minded sociologists
and society-conscious historians. See for example, Ralf Dahrendorf, Gesellschaft und
Demokratie in Deutschland (Munich, 1965), and Helmut Bohme, Deutschlands Weg
zur Grossmacht (Cologne, 1966).

The old, agonized discussions of the relationship between history and the social
sciences are increasingly irrelevant, as disciplinary lines are not merely being crossed but
trampled under foot. The arbitrary institutional distinction between history and
sociology simply disappears in monographs by Charles Tilly and Elinor Barber.
Economic theory and economic history are closely integrated by scholars such as
Douglass North, Albert Fishlow, Peter Temin, and many others. The academic
difference between political science and history vanishes in work by William Nisbet
Chambers, Samuel Beer, Stein Rokkan, Giovanni Sartori, and Hans Daalder. Demog-

FALLACIES OF COMPOSITION

219

Some historians study groups as particular events. Others are in-
creasingly interested in the regularities of group types. Many scholars
merely attempt empirical descripitions of human groups. A few construct
group patterns as paradigmatic explanation models for the resolution
of problems of individual behavior or of collective social needs.

But whatever purpose may prevail, and whatever group type may be
chosen, there is a tacit logic which attaches to all group description or
group explanation — a logic which is relevant to most historical inquiry.
And there are also many forms of illogic, which are more easily iden-
tified than the logic which they imply. The following fallacies are com-
mon cases in point.

The fallacy of composition consists in reasoning improperly
from a property of a member of a group to a property of the group itself.
This form of error is not restricted to groups of human beings but extends
to all classes of things. And as such, it occurs in two varieties: First, it
falsely extrapolates a quality of one group member to all group members.
A hypothetical example would be committed by a man who observes
that a particular American is rich and infers that all Americans, individ-
ually, are rich. Second, it is possible to transfer a quality of a member
to the group itself. The man observes one rich American and concludes
that America itself, as a group, is rich.

An actual example of the fallacy of composition appears in a book
about the American South. The subject is a tough one, much complicated
by the fact that, though there is a group of people (as well as a geo-
graphical region) which can be collectively identified as the South, there
are also many subgroups: the tidewater-Chesapeake South, the Appa-
lachian South, the Carolina low-country South, the Kentucky bluegrass
South, the Mississippi levee-and-delta South, the Alabama-Georgia black-
belt South, the Texas cow-country South, the Ozark South, several dis-
tinct Floridian Souths, and sundry others. All of these various Souths
have certain qualities in common, which mark them off from the various
Norths, but they are also unlike each other in many different ways.

W. J. Cash's Mind of the South is a classic attempt to characterize the
collective group which is the South. But its thesis is deeply flawed by the
fallacy of composition. Cash was himself a product of the Appalachian

raphy and history are one in the scholarship of Louis Henry, Etienne Gautier, Pierre
Goubert, D. E. C. Eversley, D. V. Glass, Angel Rosenblatt, and many others. All of
these works have one quality in common — they are all addressed to one or more of
the five aspects of group behavior listed in our definition above.

220

EXPLANATION

South. He was born in Gaffney, South Carolina, educated at Wofford
College and Wake Forest, and was a resident of Charlotte, North Carolina.
His perspective has been unkindly called the "Hillbilly view" of Southern
history. For Cash, "the man at the center" of Southern culture was the
man who scratched out his living on an Appalachian hillside. The in-
habitants of the other Souths are seen at a distance, like people from a
mountaintop. From that range they look small and insignificant — the
"Virginians and their artificial influence," the Charlestonians, and the
"Orleannais" who crawl around the edges of Cash's book. A critic of the
book, C. Vann Woodward, has observed that "Cash knew nothing first-
hand of the Tidewater, the Delta, the Gulf, the Blue-Grass or the Trans-
Mississippi south. But the hill country had heard quite enough of the
pretensions and poses of folk from those remote and high-living parts,
and Cash figured it was time to let them have an unbiased history of the
south from the hillbilly point of view." 4 More than that, when Cash
studied a low country nabob, he saw a hillbilly in disguise. And when he
studied a Negro he saw — he knew not what. Some strange creature whose
innards were a mystery to him, and whose presence was largely ignored.

There are other complaints to be entered against Cash's idea of the
South. There is a certain disturbing time'essness in his conceptualiza-
tions. The first half of Southern history is forgotten, and the second half
seems to float in a great fluffy cloud of eternity. The mind of Cash's South
is a singular thing, without spatial variations or temporal refinements.
Moreover, there is something of the fallacy of the counterquestion in his
book. Cash took the plantation legend and turned it upside down.
He stood Scarlett O'Hara on her head. When the crinolines billowed out
and down, there wasn't much to be seen of Scarlett's upper parts, but
there was a considerable display of her lower ones, which some inno-
cents in our own century naively persist in mistaking for reality. Scar-
lett's lower parts make a splendid spectacle. But it is a little disconcerting
to find, in a book called The Mind of the South, so little brain and so
much bottom.

Cash's book, for all its flaws, remains a very great book indeed. For
anybody who is interested in the hillbilly South, it remains the indis-
pensable guide — a veritable Baedecker to the boondocks. And anybody
who is interested in Southerners has to know about hillbilly Southerners.
We can say without condescension that Cash's Mind of the South is still
supreme in this more restricted field. But as for the rest of the South . . .
The author himself mentions Lula Vollmer's mountain woman, "who

4. C. Vann Woodward, "White Man, White Mind," The New Republic, December 9,
1967, pp. 28-30.

FALLACIES OF COMPOSITION

221

knew of France only that it was 'somers yan side of Asheville.' " Cash
may have had the same impression of Charleston, and New Orleans, and
Williamsburg, and the other distant places. He never even knew how far
away they were from his Man at the Center. 5

A second cluster of examples of the fallacy of composition appears
in the historiography of the eighteenth-century Enlightenment — an in-
ternational phenomenon if ever there was one. And yet in most accounts
there is usually a Man at the Center, as in Cash's book, or a small group
of Men at the Center, whose characteristics are generalized into the
characteristics of the Enlightment itself. Ernst Cassirer organizes his
understanding of the Enlightenment around a linear progression from
Leibnitz to Kant. The development of Kantian thought predominates.
For the English historian Alfred Cobban, the emphasis is different. "It
seems to me, on the contrary, that such thinkers as Bacon, Newton,
Locke, Hume and Bentham occupy key positions in the whole evolu-
tion of 'enlightened' Europe," he writes. Carl Becker's Enlightenment
meant Montesquieu and Voltaire and Rousseau, Diderot, Helvetius,
Holbach, Turgot, Quesnay, and Condorcet. France, he wrote, was "the
mother country and Paris the capital." J. L. Talmon's Man at the
Center is Robespierre. Lester Crocker's is, in one work, the Marquis
de Sade. There are many similar absurdities in the literature on a
subject which cries out for a careful, balanced, comparative study. 9

The fallacy of division is the converse of the fallacy of com-
position. It occurs when somebody reasons falsely from a quality of the
group to a quality of a member of the group. Once again, it applies to
all kinds of classes of things, as well as to groups of people. A hypo-
thetical example is an argument such as:

A schooner is a common type of sailing craft.
A bugeye is a schooner.

Therefore, a bugeye is a common type of sailing craft.

5. W. J. Cash, The Mind of the South (New York, 1941), p. 47. A new study of Cash
which agrees with none of these observations is Joseph L. Morrison, W . J. Cash: South-
ern Prophet (New York, 1967). I am much indebted to the insights of C. Vann
Woodward, in the article cited above.

6. Ernst Cassirer, The Philosophy of the Enlightenment, trans. Fritz C. A. Koelln
and James P. Pettegrove (Boston, 1951); Alfred Cobban, In Search of Humanity:
The Role of the Enlightenment in Modern History (New York, 1960), pp. 7-8; Carl
Becker, The Heavenly City of the Eighteenth Century Philosophers (New Haven,
1932), pp. 33-34; J. L. Talmon, The Origins of Totalitarian Democracy (New York,
1960) ; Lester G. Crocker, Age of Crisis: Man and World in Eighteenth Century French
Thought (Baltimore, 1959).

222

EXPLANATION

Both of the premises are empirically true. But the conclusion, which is
false, commits the fallacy of division.

One common form of the fallacy of division is an argument that a
quality which is shared by some or many members of a group is shared
by all of them, or that a quality which is shared by most members is
possessed by any one of them. Vernon Parrington often committed this
mistake in his great Main Currents of American Thought, where he
formed the habit of conceptualizing a problem in group stereotypes and
transferring the stereotype to individual members. Thus,

Most Calvinists were theological determinists.
Most New England Puritans were Calvinists.
Therefore, most New England Puritans were theological
determinists.

But if a good deal of recent scholarship is right, New England Puritans
were not determinists, or at least they were determinists with a differ-
ence.

A second type of the fallacy of division is an inference from a
property of the group itself to a property of the member of the group.
Thus, again from Parrington (2:179),

The fortunes of the Federalists decayed after 1800.
Joseph Dennie was a Federalist.

Therefore, the fortunes of Joseph Dennie decayed after 1 800.
There is only one remedy for this form of error — research.

The fallacy of difference consists in a tendency to conceptual-
ize a group in terms of its special characteristics to the exclusion of its
generic characteristics. 7 This method is fallacious if the line of distinction
between special and generic qualities cuts across the line of inquiry — as
it almost invariably does. The group called Puritans serves as a useful
example. Only a small part of Puritan theology was Puritan in a special
sense. Much of it was Anglican, and more was Protestant, and most was
Christian. And yet Puritanism is often identified and understood in terms
of what was specially or uniquely Puritan.

The fallacy of difference is often an attempt at a definition by genus
and difference, in which the genus is omitted or forgotten. "Such a reduc-
tion," writes Ralph Barton Perry, "is not only narrow, but false; because

7. It is identified and labeled by Ralph Barton Perry, in Puritanism and Democracy
(New York, 1964), p. 82.

FALLACIES OF COMPOSITION 223

the differences lose their meaning when divorced from the genus." 8 A
specimen is John F. H. New's Anglican and Puritan: The Basis of Their
Opposition, 1558-1640 (Stanford, 1964). If New merely meant to
establish the points of conflict between Puritanism and Anglicanism, there
would be nothing fallacious in his approach. Everything, as ever, depends
upon purpose. But he moves beyond this purpose to another, for which
his procedure is fallacious and seriously inconsistent — an attempt to
generalize about the theology and psychology of Puritanism, which he
undertakes to do in a rounded way but accomplishes only in terms of the
points of conflict between Anglicans and Puritans. Many common com-
mitments which appeared in both Anglicanism and Puritanism, however,
are relevant to this larger purpose. In many respects, most Puritans were
Anglicans. They were not Anglican in the same way as Archbishop Laud,
but neither were they antithetical in all respects.

Another example of the fallacy of difference is identified by Bar-
rington Moore, in the historiography — and sociology — of India, with
respect to the significance of the caste system. "At least in its full ramifica-
tions," he writes, "the caste system is unique to Indian civilization. For
this reason there is a strong temptation to use caste as an explanation for
everything else that seems distinctive in Indian society. Obviously this
will not do." Moore notes that caste was used by historians to explain
the apparent absence of religious warfare in India, but more recently
religious warfare has become a problem of major proportions in India,
and the caste system remains. I suspect that the caste system has been
exaggerated not merely as an explanation of other distinctive aspects of
Indian civilization but as an explanation of characteristics which were
not distinctive as well.

The converse fallacy of difference, on the other hand, renders
a special judgment upon a group for a quality which is not special to it.
Commonly, this error appears in condemnatory judgments. Once again,
the historiography of Puritanism provides an example. In America today,
it is probable that Puritans are better known for the burning of witches
than for anything else (except, perhaps, the baking of turkeys).

But by the standards of their own time, New England Puritans ap-
pear to have been extraordinary for the fact that they killed few witches
and burned none. The incidence of executions for witchcraft seems to
have been lower, per capita, in New England than in Old England, and

8. Ibid., p. 82.

9. Barrington Moore, Social Origins of Dictatorship and Democracy (Boston, 1966),
p. 334.

224

EXPLANATION

much lower in America than in the countries of continental Europe.
Estimates by able historians vary from a low of several hundred thousand
executions for witchcraft in Western Europe in the early modern period
to a high of several million. The few dozen murders in New England were
modest by comparison. None of this, of course, is meant to condone witch-
craft executions or to vindicate Puritans. But it is meant to suggest that
there was nothing specially Puritan about witch hunts, and that there was
possibly an inverse correlation between the two phenomena. 10

A comparable error also appears in the history of the South. John
Hope Franklin, in a monograph called The Militant South (Cambridge,
Mass., 1956), argued that Southern culture was pervaded by violence.
Few scholars would, I think, challenge that conclusion. But Franklin
also suggests that violence was Southern in a special sense and that it was
a particular consequence of the South's Peculiar Institution. The infer-
ence is surely mistaken. Much of what Franklin found might be described
as a Western as well as a Southern phenomenon. And in the cities of the
Northeastern states, violence in the streets during the period 1800-1861
was a common and ordinary fact of life. Baltimore was not the only city
that deserved the name of "Mob-town." There were many ugly and bloody
disturbances in the streets of New York, Philadelphia, and even Boston.

There was also much violence in the Old Northwest. A history of
one Illinois county is suitably titled Bloody Williamson. Franklin's study
of violence is not quantitative — nor is anybody else's in this period. Until
a controlled investigation is completed, it appears exceedingly unlikely
that a high incidence of violence was uniquely Southern in the period
1800-1861. Maybe there was something specially Southern about the
quality of violence below the Mason-Dixon line. And there is some
evidence that violence may have been institutionalized in the South in
ways which were not apparent in the Northeast or Northwest — in the
establishment of military schools for example, and in the ritualized
violence of the Southern elite. But in its impressionistic conclusions as
to the incidence of violent acts, Franklin's Militant South seems to be in
error.

$»» The fallacy of ethnomorphism is a form of error which is
worth distinguishing from ethnocentrism for analytical purposes, though
the two are often compounded in a single name, and sometimes in a
single act. Ethnomorphism is the conceptualization of the characteristics

10. George L. Kittredge, Witchcraft in Old and New England (Cambridge, Mass.,
1929); H. R. Trevor-Roper, Religion, the Reformation, and Social Change (London,
1967), pp. 90-192.

FALLACIES OF COMPOSITION

225

of another group in terms of one's own. This pernicious error was
recognized and repudiated two hundred years ago, but a perceptive
English historical critic, Bolingbroke, complained that "There is scarcely
any folly or vice more epidemical among the sons of man, than that
ridiculous and hurtful vanity by which the people of each country are
apt to prefer themselves to those of every other, and to make their own
customs and manners and opinions the standards of right and wrong, of
true and false." 11

In historiography, the same error appears not merely in moral
judgments but also in behavioral understandings. It is committed by
primitive and civilized people alike. When the Puritans began to settle
in Massachusetts, their Indian neighbors watched them move in and
wondered why they had come. The Narraganset finally decided that the
English must have burned up all the firewood in the old country and had
moved to find more. "This was one of the Indians' chief reasons for re-
moval," a historian explains, "and they naturally projected it to the
white man." 12 The Puritans, in turn, pondered the presence of the red
man in America, and some persuaded themselves that the Indians were
descendants of the ten lost tribes of Israel — a historical interpretation
which retained its popularity in Protestant America for the next two cen-
turies. Cotton Mather had a slightly different thesis. "Though we know not
when or how these indians first became inhabitants of this mighty con-
tinent," he wrote, "yet we may guess that probably the devil decoyed those
miserable savages hither, in hopes that the Gospel of the Lord Jesus
Christ would never come here to destroy or disturb his absolute empire
over them." 13

There are many modern examples. In ancient historiography, Gil-
bert Murray once asserted that Homeric religion was "not a religion at
all." He meant that it did not meet the descriptive definitions of religion
in modern Western culture. 14 Similarly, Lewis Mumford has complained
of ethnomorphic conceptions in modern interpretations of the history of
ancient technology, which sometimes reads like the chapter from The
Natural History of Iceland that Dr. Johnson discovered — a chapter which
which was called "Concerning Snakes," and read in its entirety: "There
are no snakes to be met with throughout the whole island." 13 Mumford
observed,

11. Henry Bolingbroke, Letters on the Study and Use of History (London, 1870),
pp. 9-10.

12. Alden Vaughan, The New England Frontier (Boston, 1965), p. 62.

13. Ibid., p. 20.

14. Gilbert Murray, Tradition and Design in the Iliad, p. 222.

15. James Boswell, Life of Dr. Johnson, Modern Library ed. (New York, n.d.,),
p. 799.

226

EXPLANATION

What has misled judgment in our own age is that the greatest technological
achievements of the ancient world were in the realm of statics, not dynamics,
in civil, not mechanical engineering: in buildings, not machines. If the his-
torian finds a lack of invention in earlier cultures, it is because he persists in
taking as the main criterion of mechanical progress the special kinds of
power-driven machine or automation to which western man has now com-
mitted himself, while treating as negligible important inventions, like central
heating and flush toilets — or even ignorantly attributing the latter to our own
"industrial revolution." 16

Other examples of ethnomorphism appear in broad conceptions of
historical development. To George Bancroft it seemed that all groups
would gradually develop into the form of the American Republic
circa 1840. To Leopold Von Ranke, the model was the Prussian
monarchy. Each scholar made the mistake of assuming that the develop-
ment of his own particular group was a prototype for the development of
all groups.

There is a vast ethnomorphism in Freud's thought, in the form of
assumptions that behavioral patterns which appeared in the members
of his own culture were transf errable without major change to members of
many other groups, far removed in space and time. When Freud himself
did this in his analyses of Leonardo and Moses, the result was an inter-
pretative atrocity. Similarly, Marx generalized certain group characteris-
tics which did exist (some of them) in nineteenth-century England,
France, and Germany into universal phenomena for all groups every-
where.

Today the most powerful form of ethnomorphism is the idea that
Anglo-American and North European cultural characteristics are the
cultural norm. This mistaken belief is not merely prevalent among Anglo-
Americans and North Europeans. The painful zeal with which Negro
ladies have their hair straightened, Jewish ladies have their noses
straightened, and Oriental ladies have their eyes straightened suggests
that this pervasive ethnomorphism is internalized by members of other
groups. Reference-group theory helps us to understand that many human
beings have arranged their lives with reference to groups to which they
do not belong. 17

$»» The fallacy of ethnocentrism is committed by a historian who
exaggerates the role of his own group in its interaction with other groups.
A historian may commit it by mistaking not the shape of another group

16. Lewis Mumford, The Myth of the Machine (New York, 1967), p. 244.

17. Herbert H. Hyman and Eleanor Singer, eds., Readings in Reference Group Theory
and Research (New York, 1968).

FALLACIES OF COMPOSITION

227

but merely its size — much as the endpaper map in Cleveland Amory's
Proper Bostonians deliberately changes the size of American states in
proportion to the ethnocentric perceptions of the inhabitants of Boston.

A historiographical example is provided by Chester Wilmot, a
British journalist who published an excellent military history of. World
War II. He made a serious mistake, however, in his suggestion that the
Russo-German War was an extension and a consequence of a prior
Anglo-German rivalry:

This was Hitler's solution of the dilemma in which he was placed by Britain's
refusal to yield. The essence [that word!] of his problem was that he could
not gain the resources for a prolonged struggle with the Anglo-Saxon powers
without involving himself in that "war on two fronts" which he had sworn
to avoid. He could not inflict a crippling defeat on the British, let alone the
Anglo-American combination, until he had greatly expanded his Navy and
Air Force and was free to concentrate the main strength of his Army against
the West. This he could not do so long as the threat of hostile Soviet action
in the East compelled him to divide his forces and to allocate to the army
two-thirds of his mobilised manpower and of his armament production.
Furthermore, he could not keep Occupied Europe fed and quiet during the
conquest of Britain, nor could he wage the intensive air and naval warfare
which alone could bring victory, unless he could be certain of two things:
full control over the economy of the Balkans, and continued deliveries of
grain and oil from the U.S.S.R. 18

An opposite interpretation might be nearer the mark. Hitler's career was
more clearly marked by an obsession with Eastern Europe and Bolshe-
vist Russia, rather than by a prior determination to conquer England.
Along the way, he appears to have found himself in the unwelcome
position of being compelled to fight England, and not the other way
around. Moreover, Wilmot's map of the struggle magnifies the size of
the island of Britain, much as Mercator's projection enlarges the island
of Greenland. At the same time, it shrinks the great land masses of
Russia and the United States.

On the other side, the Russian version of the Second World War
errs in the opposite direction, and in greater degree. In the U.S.S.R. the
Second World War is known as the Great Patriotic War. The official
multivolumed history provides factually inaccurate estimates of the num-
ber of German troops engaged in the East and in the West. And the world
to the west is shrunk to the size of a Leningrad suburb. American authors,
also, have made comparable mistakes in standard histories of the French
and Indian War, the War for Independence, the Wars Against African
Piracy, the War of 1812, World War I, and World War II. In all of

18. Chester Wilmot, The Struggle for Europe (New York, 1952), p. 57.

228

EXPLANATION

these instances, the importance of American groups is grossly exagger-
ated, and the roles of other groups are much diminished.

There are many other examples in military history. Field Marshal
Slim once denned a battle as something which happens at the intersection
of two or more maps. We might paraphrase his statement: a battle is some-
thing which is fought at the intersection of two or more cultural groups.
The temptations of ethnocentrism in this situation are almost irresist-
ible.

But military history is not unique in this respect. The same fallacy
appears in social, cultural, and political history. A striking example is
the historiography of modern Burma, which calls to mind the Hindustani
fable of the Six Men and the Elephant. Many different national groups
interacted in the history of Burma; besides the Burmese, there were
Chinese, Indians, British, Americans, Frenchmen, Japanese, and others.
Fine Anglocentric books about Burma have been written by Maurice
Collis, John S. Furnivall, D. G. E. Hall, and G. E. Harvey, to name
but a few of many English and Anglo-Burman authors. Other scholars
have produced works on the same subject from an American perspective,
notably John L. Christian and John F. Cady — works which tend perhaps
to overemphasize the admittedly important role of American missionaries
in Burma. There are a few histories of Burma from the Japanese point
of view — Willard Elsbree's Japan's Role in Southeast Asia (Cambridge,

1953) . Still other accounts are Sinocentric, Indocentric, or Francocen-
tric. Each of these approaches tends to exaggerate the role of a particular
ethnic group in a very complex pattern of multiethnic interaction.

We have heard from each of the Six Men in Burmese historiography
— in this case an Englishman, an American, an Indian, a Japanese, a
Chinese, and a Frenchman — all of whom tended to exaggerate the size
and significance of that part of the elephant upon which his hands hap-
pened to rest. But recently we have begun to hear from the elephant.
Histories of Burma by Burmese scholars and statesmen are beginning to
appear in quantity. 19 These works are, if anything, more stridently
ethnocentric than those which preceded them. They are painful works
of pious devotion to the Burmese people — works comparable in spirit
to the heroic acts of Burmese ladies of good breeding before the British

19. Maung Htin Aung, A History of Burma (New York, 1967); Maung Maung,
Aung San of Burma (The Hague, 1961); U Nu, Burma Under Japanese Rule (London,

1954) ; Ba Maw, Breakthrough in Burma: Memoirs of a Revolution, 1939-1946 (New
Haven, 1968); U Ba U, My Burma (New York, 1959). For a review of earlier historical
writing in Burma see Thaung Blackmore, "Burmese Historical Scholarship," in
Historical, Archaeological and Linguistic Studies on Southern China, Southeast Asia
and the Hong Kong Region (Hong Kong, 1967), pp. 310-20.

FALLACIES OF COMPOSITION

229

"intrusion," as Maung Htin Aung calls it (p. 210) — ladies who, "as an
act of piety to obtain merit, volunteered to suckle infant elephants,"
which were the sacred symbols of Burmese culture and religion. 20 There
is a good deal of vicarious elephant suckling in Burmese national history.
But it is surely no worse than the ethnocentric effusions of historians in
other nations.

There are other studies of Burmese history and culture which are
remarkable for their comparative absence of ethnocentrism. A very great
example is E. R. Leach's Political Systems of Highland Burma (London,
1954), which is sufficient evidence that this bias can be controlled.
Leach has written of group interactions which are conceptualized in such
a way as to neutralize his own ethnocentric bias. The same effect can be
achieved in many contexts, with ingenuity and common honesty.

The problem which appears in Burmese historiography commonly
occurs in much more aggravated forms elsewhere. A student of In-
donesian history has observed with respect to this field that

the fundamental historiographical problem is, then, to find the meeting
point between the many local histories of the Indonesian people and colonial
history and so to determine how to unite the two. What criteria are to be
used in order to write a single narrative out of so many histories? Is it pos-
sible to blend a history of Indonesia which is Indonesian in character (In-
donesia-centric) out of local histories which are regie-centric or ethnocentric
and out of colonial histories which are to us essentially [that word again!]
xenocentric in character? 21

These are, indeed, hard questions. When they are cast in these terms,
they are perhaps insoluble. Though an "Indonesia-centric" history of
Indonesia may seem at first sight to be perfectly sensible and defensible,
and though it is surely functional to nationalizing purposes, it is dysfunc-
tional to sound scholarship. Indonesian history has been made, in an
extraordinary degree, by the interaction of many different groups, cul-
tures, and nations. The only defensible approach, and the only effective
solution to the historiographical problem, is a balanced and refined
polycentrism, difficult though that may be to sustain.

Another example of ethnocentrism appears in European historiog-
raphy. A French historian of England, Elie Halevy, placed heavy
emphasis upon the French Revolution of 1830 in his interpretation of
the progress of English reform during the 1830s. 22 An English historian,
Norman Gash, has made a persuasive argument for the proposition that

20. John F. Cady, A History of Modern Burma (Ithaca, N.Y., 1958), p. 7 n.

21. Mohammed AH, "Historiographical Problems," in Soedjatmoko, ed., An Introduction
to Indonesian Historiography (Ithaca, 1965), pp. 11-12.

22. Halevy, The Triumph of Reform (1830-1841) (London, 1961), pp. 3, 34-37,
passim.

230

EXPLANATION

Halevy exaggerated the importance of French affairs in English politics,
and particularly in the English General Election of 1830. 23 All Gaul may
be divided into as many parts as one likes, but one of them is not England.
Halevy's error (if error it is) is an aberration in a work which is re-
markably clean of ethnocentric fallacies — and an extraordinary model
of intercultural understanding.

The fallacy of elitism consists in conceptualizing human
groups in terms of their upper strata, or of casting belief units in terms of
their most refined thoughts and elegant expressions. Elitism substitutes
Society for society, Culture for culture, Civilization for civilization, Man-
ners for manners, and Morals for morals. This semantical ambiguity
appears in many languages. The terms "Civilization," "Civilisation,"
and "Civilization" are involved in the same contradictions in English,
French, and Spanish. Dante used "Civilta" in the same ambiguous way.
And the Germans have assimilated a splendid variety of pedantical
terms— "Kultur," "Civilisation," "Gesittung," "Bildung," "Ausbildung,"
"Sitte," "Gewohnheiten," "Gebrauche" — most of which are equally
treacherous.

There is, of course, no impropriety in a study of Proper People, or
of their Best Thoughts. A history of Society can be quite as useful as a
history of society. But there is a fallacy in a confusion of the two, which
has often happened in historical writing. The results sometimes include
serious substantive error, as well as conceptual superficiality. In the
Bosporus there is a surface current which flows powerfully to the south-
west. But beneath it the water speeds swiftly in the opposite direction.
The same phenomenon sometimes appears in human groups.

The fallacy of elitism appears in a great pioneer work of cultural
history, Voltaire's Age of Louis XIV. Voltaire's interest in manners and
customs constituted an important and constructive departure from
political and dynastic history. But the transition was incomplete. In
place of a history of a political dynasty, Voltaire wrote the history of a
Cultural dynasty. Worse, he seems to have assumed that where there
was no Culture, there was no culture. In his preface he wrote:

The thinking man, and what is still rarer, the man of taste, numbers only
four ages in the history of the world; four happy ages when the arts were
brought to perfection and which, marking an era of the greatness of the
human mind, are an example to posterity.

23. Norman Gash, "English Reform and the French Revolution in the General
Election of 1830," Pares and Taylor, eds., Essays Presented to Sir Lewis Namier
(London, 1956), pp. 258-64.

FALLACIES OF COMPOSITION

231

The first of these ages, to which true glory belongs, is that of Philip and
Alexander, or rather of Pericles, Demosthenes, Aristotle, Plato, Apelles,
Phidias, Praxiteles; and this honour was confined within the limits of Greece,
the rest of the known world being in a barbarous state.

The second age is that of Caesar and Augustus, distinguished moreover
by the names of Lucretius, Cicero, Livy, Virgil, Horace, Ovid, Varro and
Vitruvius.

The third is that which followed the taking of Constantinople by
Mahomet II . . . the hour of Italy's glory. . . .

The fourth age is that which we call the age of Louis XIV, and it is per-
haps of the four the one which most nearly approaches perfection. Enriched
with the discoveries of the other three it accomplished in certain departments
more than the three together. All the arts, it is true, did not progress further
than they did under the Medici, under Augustus or under Alexander; but
human reason in general was brought to perfection. 24

This statement is a little imperfect in its chronology and a little
credulous in its ascription of greatness. How can Apelles be an example
to posterity? None of his works is known to survive. But there are
more serious objections. Always, as Voltaire himself privately said of the
book, "the principal figures are in the foreground; the crowd is in the
background." Everywhere the object was to celebrate genius and cultiva-
tion. If Culture did not exist, in Voltaire's narrow conception, then there
was nothing. Cultural history became a learned society — a club with a
very exclusive admissions policy. If the subject matter of Voltaire's work
was different from that of so-called drum-and-trumpet history, the ruffles
and flourishes still sound the same.

Voltaire's confusion of Cultural history with cultural history still
appears in modern historical works by such so-called cultural historians
as Jacques Barzun, to name but one. Barzun writes about great musicians
the way drum-and-trumpeters celebrated great generals. To brass and
percussion he adds strings and winds. But everything else is still the
same. He professes cultural history, but produces Cultural history. Many
cultural histories are still Cultural in their content. Examples are two
recent volumes in Harper & Row's New American Nation series — Louis
B. Wright's The Cultural Life of the American Colonies (New York,
1957) and Russel B. Nye's The Cultural Life of the New Nation (New
York, 1960). Both books begin with prefatory promises of cultural his-
tory. But they describe Culture instead — a weary round of stale anecdotes
and impressions, mostly about the Best People.

Another example is the writing of a Latin American historian,
Jose Toribio Medina. His work excluded the lower classes, "from whom

24. Voltaire, The Age of Louis XIV, trans. Martyn Pollack. Everyman edition (London,
1961), p. 1.

232

EXPLANATION

we can learn nothing," he wrote. 25 It is something of an irony that this
scholar, who so cavalierly dismissed the vast majority of humanity from
history, should be celebrated by a band of admirers as the "Humanist of
the Americas." 28

The fallacy of racism is a popular delusion, and all the more
powerful for its tendency, increasingly, to run underground. It might be
defined in three different ways:

1. A false classification of people into fixed biological groups.

2. A false explanation of culturally learned behavior in terms of a
biological, physiological, or hereditary cause.

3. A false prejudice, for or against any genetic class or ethnic group
of human beings.

These three forms of error commonly coexist. But any one of them alone
is sufficient to constitute the fallacy of racism.

Everything about this subject is controversial. But some things seem
reasonably clear. Thanks to the revolutionary progress of genetics in the
past generation, it can be demonstrated beyond a reasonable doubt that
there are classes of men which can be called races — classes which are
most accurately called "gene pools," or "breeding populations." A race
is a collection of people who share a common genetic heritage and certain
statistical regularities in their genetic make-up, which distinguish them,
in some degree, from other collectivities. 27

But, having asserted this, we must quickly add three important

25. Arthur P. Whitaker, "Medina's Concept of History," in Maury A. Bromsen, ed.,
Jose Toribio Medina, Humanist of the Americas (Washington, D.C., 1960), p. 70.

26. There are many other examples. Prominent among them is Werner Jaeger,
Paideia: The Ideals of Greek Culture, 3 vols., trans. Gilbert Highet (New York, 1936) —
a heavily elitist view of Greek culture. Cf. Moses Hadas, Hellenistic Culture: Fusion
and Difference.

27. I am following two moderate interpretations of race, Stanley Garn, Human
Races (Springfield, III., 1961), and Theodosius Dobzhansky, Heredity and the Nature
of Man (New York, 1964). There is a wide range of expert opinion. Some anthropolo-
gists and geneticists have, I think, allowed too little latitude to race. Prominent among
them is M. F. Ashley Montagu, who comes very close to denying that race exists, in
Man's Most Dangerous Myth: The Fallacy of Race, 3d ed. (New York, 1952).
Montagu has also edited a collection of essays in the same vein by ten scholars, all of
whom minimize race, in The Concept of Race (New York, 1964). On the other
hand, Carleton S. Coon, in The Origin of Races (New York, 1962) and The Living
Races of Mart (New York, 1965), argues beyond his evidence, in an interpretation
which allows too much latitude to race differences. Between Ashley Montagu and
Carleton Coon, there is a mediating position — the position of Garn and Dobzhansky,
if I understand them correctly — which is in my judgment more tenable than either
extreme, in light of present evidence.

FALLACIES OF COMPOSITION

233

qualifications. First, there is no simple definitive taxonomy of these
"breeding populations" — no clear and crisp set of classes. There are
many different variations in the genetic make-up of men, and there is no
easy way of sorting them out. Some careful students of race divide men
into a mere handful of racial groups; others, into several hundred. 28 For
Carleton Coon, there are five races in the world — Caucasoid, Mongoloid,
Congoloid, Australoid, and Capoid, 28 with numerous "clines" or inter-
racial shadings between. Other scholars distinguish many small and
special "races," even such as "Pitcairn Islanders" and "Neo-Hawai-
ians." 30 Wherever there has been a group which has tended to inter-
marry, there is likely to be a gene pool, which can be defined with respect
to some particular genetic peculiarity. Jews appear to be genetically
distinct with respect to a particular statistical predisposition to certain
rare hereditary diseases, such as the Tay-Sachs disease, and familial
dysautonomia. Ashkenazic Jews are genetically distinguishable from
Sephardic Jews with respect to familial Mediterranean fever and favism.
The Irish appear to have a monopoly on the rare hereditary disease
called leprechaunism. Scandinavians are apparently unique in the inci-
dence of Silferskiold's disease. But none of these groups are now known
to possess sufficient genetic differences to warrant special identity as a
"race." 31

Second, in the past five hundred years breeding populations have
been highly unstable. Many new gene pools have been formed; others
have faded away. Research in biochemistry and genetics has demon-
strated that races have been in process of rapid and complex genetic
change — more rapid and complex than most scholars believed possible a
generation ago. 32

Third, races are not "groups" in our sense of the word. They do,
beyond a doubt, meet two of our group criteria, but not others. Races do
possess a finite membership which can be statistically specified. And
they do possess a common history. But they have no group structure, and
no group function, and probably no normative patterns of conduct.

It is readily apparent that race is more than skin-deep. Gam ob-
serves that "racial differences are known to exist in almost every area
of anatomy where comparative data have been accumulated, and there

28. Garn, pp. 12-38, 116-32.

29. Coon, The Living Races of Man, pp. 6-10, passim.

30. Garn, pp. 7, 132. There has even been a classification of classifiers into "lumpers"
and "splitters" — the former preferring a few big categories, and the latter, many
little ones (ibid., p. 12). No genetic differences between lumpers and splitters have
been demonstrated.

31. Ibid., pp. 81-91.

32. Ibid., p. v.

234

EXPLANATION

is growing evidence for racial differences in biochemical functioning and
in the constituents of cells and tissues." 33 Racial differences have been
located with some precision in pigmentation, hair form and color and
quantity, bone size and form, tooth structure, growth rates, blood groups,
susceptibility to disease, adaptations to heat and cold, sensitivity to drugs,
and even in the composition of earwax.

These physiological variations are established beyond cavil. More
doubtful is the problem of racial patterns of temperament and intelli-
gence. No such patterns have been proved to exist. But this is not to say
that such patterns have been proved not to exist. Gam writes,

Racial differences in measured intelligence thus remain neither proven nor
disproved. There are differences, but like stature, they do not necessarily in-
dicate the maximum level of capacity in the absence of standard or con-
trolled conditions. To the confirmed believer in racial differences in intel-
ligence, we can simply say that the more nearly two groups are matched in
educational level, family background, opportunity and security, the closer
they agree on averaged I.Q. scores. To the dedicated equalitarian, the be-
liever in no race differences, the disparate levels in the currently best-matched
Negro-white comparisons remain to be refuted. ... A very reasonable guess
is that races are comparable in the sum and total of what we call "intel-
ligence," but differ in many interesting details. As with the automatic re-
sponse patterns that so neatly differentiate one individual from the other,
race differences may exist in form-discrimination, color-sense, tonal-memory,
mechanical-reasoning, abstract-reasoning and with other special (rather than
general) aspects of intelligence. This supposition, moreover, is directly sus-
ceptible to testing. 34

It is, indeed, susceptible to testing, but as yet it has not been tested
in a conclusive way. The question must, at this moment, remain unan-
swered. An open-minded historian will, I think, carefully follow the
progress of biochemical genetics, which has already produced some find-
ings which are directly relevant to major historical problems. The genetic
adaptation of Negroes to humid heat has been demonstrated by careful
research, as well as the so-called sickle-cell trait, which provides a defense
against malaria. Both of these facts cannot be ignored by a historian of
the institution of Negro slavery. They assume great importance in any
study of the history of the Black Caribs, for example. It is likely that
other genetic discoveries — possibly even with respect to intelligence and
temperament — will prove relevant to many other historical problems.

But at the same time, a historian must take the full measure of
much research in the past fifty years, which has demonstrated the extra-
ordinary cultural malleability of men — all men. The famous alpha tests

33. Ibid., p. 37.

34. Garn, Human Races, pp. 110-15.

FALLACIES OF COMPOSITION

235

of the United States Army in World War I found that Negroes from
some Northern states scored higher in one kind of intelligence than
whites from some Southern states. 35 Many recent projects have produced
similar results.

It is also important to guard against common errors of inference.
First, genetic patterns are statistical in nature. One cannot assume that
any given member of a breeding population, even if his membership can
be clearly established, will possess the genetic make-up which is normal
in his group. Second, one cannot conclude that a breeding population
which is distinctive in some ways is therefore distinctive in all ways,
or in any particular way which is not clearly and conclusively established.
Third, the arbitrary nature of all taxonomies of race should be clearly
recognized. If genetic patterns are sometimes useful to historical research,
they must be carefully defined ad hoc and not converted into sweeping
generalities about tendencies of all people of a given genetic class in all
respects. Genetic classifications must not be misconstrued as fixed groups.
Fourth, a level of performance is not a measure of capability. Fifth,
genetic classes must not be confused with social and cultural groups.
Historians have mistaken language groups such as Aryans and Semites
for racial classes. Often, in the nineteenth century, they mistook national
groups for races. Henry Cabot Lodge wrote frequently of the "English"
race. There are laws in some American states which still speak of the
"Chinese" race. An Italian novelist, Ignazio Silone, described in Bread
and Wine the tendency of some people to confuse occupational groups
with race:

Don Paolo was suprised to observe the role that mustaches, beards, and hair
still played in differentiating the professional class from the peasants and the
landlords. He realized also why the various classes were indicated in dialect
by the word "race" — the "race" of husbandmen, the "race" of "artists"
• (Artisans), the "race" of landowners. The son of a petty landowner who
studies, and therefore inevitably becomes a state or municipal employee,
promptly tries to obliterate the fact that he comes of the "race" of husband-
men by brushing his hair in the style of his new station. 36

There are many comparable errors in historical scholarship.

Finally, a historian should beware of the counterfallacy of anti-
racism, which has been committed by many well-meaning social scien-
tists. "Racist" is today a pejorative term, as negative in its connotation as
"nigger" used to be — if not more so. "Nigger" connoted incapacity;

35. Ashley Montagu, "Intelligence of Northern Negroes and Southern Whites in the
First World War," American Journal of Psychology 68 (1945): 161-88.

36. Ignazio Silone, Bread and Wine (New York, 1946), p. 151; quoted in Montagu,
Man's Most Dangerous Myth, pp. 83-84.

236

EXPLANATION

"racist" suggests depravity. For this remarkable change in meaning,
within the span of a generation, Nazism may be primarily responsible.
The discrediting of racism may, paradoxically, prove to be Hitler's most
enduring accomplishment. But some people have reacted in a mistaken
way to the terrible crimes and atrocities which were committed in the name
of race. They have tended to repudiate the idea of all hereditary charac-
teristics, and to dismiss race as a dangerous superstition. Some racist
thought is dangerous, but genetics remains a science. Race itself, properly
understood, is a reality — a historical reality — at the same time that
racism is a profound and bloody error, which cannot be tolerated in the
contemporary world. Men can live with other fallacies in this book. But
racism kills and maims and mutilates. Against this form of error, there
ought to be a law.

The fallacy of cross grouping consists in the conceptualization
of one group type in terms of another. Hypothetically, this form of error
can occur in numerous ways — in n 2 ways, if n = the number of group
types in the world, plus innumerable multiple combinations. Practically,
however, historiographical cross-grouping tends to appear in one partic-
ular variety: the conceptualization of all group types in terms of the na-
tion-state. There are, of course many useful particular problems which
can and should be conceptualized in terms of the nation-state — mostly
political and legal problems, for by definition a nation-state is a legal
and political group. But there are many other problems which should
be approached differently — problems about religious, economic, social,
or cultural groups, which rarely coincide with the nation-state. It would
make no sense to write a history of the German religion, the Swiss lan-
guage, the Congolese society, or the Nigerian culture. There are no such
things. There are religions in Germany, languages in Switzerland, so-
cieties in the Congo, and cultures in Nigeria. But these problems are not
defined accurately by national frontier posts. They have boundaries of
their own.

The preponderant majority of the 134 sovereign nation-states which
exist at the time of this writing do not possess a clearly defined national
culture, a single language, a national society, or a national economy.
Most nation-states in the world today are extensions of the colonial ad-
ministrations which preceded them, and they still retain much of the
colonial regime's shape and substance. A few were physically formed by
some forgotten European bureaucrat, thousands of miles away, who
scratched lines upon the map of a continent which he had never seen, nor
ever wish*d to.

FALLACIES OF COMPOSITION

237

Though national history is increasingly unfashionable nowadays, it
still retains its hold upon historians. Historical records have been sorted
into national piles; there is a bias in our source material which is not
easily overcome. Many classic historiographic problems are problems of
national history, not readily translated into other terms. In American
universities, historians are still hired as specialists in national history, and
they tend to be trained in that way as well. This is not commonly the case
in ancient or medieval historiography, and not consistently the case in
African, Asian, and Latin American historiography. But most historians
in America are specialists in modern European and United States history,
and as such they tend to be national historians.

Today, "cultural history" is replacing national history in much of
the best historical scholarship. And a "culture" is often understood to
mean an integrated group having a common political organization, a
social structure, an economic system, a religion, and a common way of
life that embraces every aspect of human existence. In this sense, "cul-
ture" is an abstraction, and an exceedingly dangerous one. Except,
perhaps, on a few remote and primitive islands, it does not exist.

But historians are tending to graft this abstraction called culture upon
the root of national history, with deleterious consequences for empirical
research and with many classic examples of the fallacy of cross-grouping.
There are some wise words in David Potter's essay, "The Historian's Use
of Nationalism and Vice Versa," 37 which identifies the common tendency
of historians "to assume too simple an equation between nationality and
culture. . . . Many 'nationalist' movements have a minimum of common
cultural content ... the impulse moving them is primarily a negative
political reaction against an existing regime (especially a colonial
regime)." 38

This modern tendency has lent new life to an old error. The nation-

37. American Historical Review 67 (1962): 924-50.

38. Ibid., pp. 933-35. The problem of nationality in historical writing is complicated
by the incorrigible habit which many historians have formed, of confusing different,
and sometimes antithetical meanings and purposes in the use of that term:

1. A confusion of "nation" as an idea, with "nation" as a group, with "nation"
as an institution. These meanings can coexist in some contexts, as Mr. Potter demon-
strates, but they are sometimes contradictory.

2. A confusion of what is with what ought to be: Mr. Potter notes the tendency
of "the historian to deny nationality to groups of whom he morally disapproves."

3. A confusion of what exists historically with what is thought to exist by
historians. Potter himself makes this mistake when he writes, "National groups usually
coincide with a political state, but it would be too restrictive to say that a national
group is simply a political group, for very often the historian is not concerned with
the political aspects of the history of the group" (p. 924). This is a lovely example of a
non sequitur. Things that are, are not the same as things that are of concern to
historians. Historiographies! interest is not an index of historical reality.

238

EXPLANATION

state, in the new form of national culture, retains its power as the
predominant unit of historiographical inquiry. Just how powerful na-
tional history still is appears in two recent attempts to escape from it.
One is by Cyril Black, who in an important new book called The
Dynamics of Modernization (New York, 1966) attempted to break the
lockstep of national history by studying in a comparative way the
phenomenon of modernization everywhere in the world. But the unit of
his analysis is still the nation-state. In the words of a reviewer of the book,
"he accepts the concept of the national state, which he identifies with
politically organized societies as almost the only really significant classi-
fication of mankind, and this leads him to some rather odd taxonomies
both of processes of development and of national states themselves." 39
Black, in the usual way, grafts onto the main stem of the nation-
state an economy, a society, and a culture. But always the nation-state
retains its pre-eminence. The reviewer notes, "To select the national
state as the unit of study in the process of modernization introduces a
serious bias into the account at the very start. It makes the whole process
look much more political and much more planned than it was." His re-
viewer, an economist, would have preferred, if I understand him, to make
ecological groups into the primary units of analysis, which is equally
mistaken. A phenomenon such as modernization, as it is usually con-
ceived, involves political development, economic development, social
development, and cultural development on many different axes. What is
required, surely, is a more complex pattern of conceptualization to deal
with this complex subject — a conceptualization in terms of a variety of
group types, which conforms more closely to the reality of multigroup
membership in the modern world. Maybe the political, economic, social,
and intellectual categories are not the most suitable ones — they are
merely a reflection of the organization of academic society. The point
is not that these particular groups should become the units of analysis,
but rather that the analytical strategy must be adjusted to the subject,
and in the study of a polycentric phenomenon, a conceptual polycentrism
is required. The complex cannot be reduced, for convenience, to the
national state, or the ecological system, or any other monistic scheme.

A second attempt to break out of the narrow boundaries of national
history is R. R. Palmer's The Age of Democratic Revolution (2 vols.
[Princeton, 1959-1964]), an excellent and useful book, which attempts
to conceptualize revolutionary change in the Western world in the period
1760-1800 in terms of a world revolution in which many nations
participated. But the transition from national to supernational history is

39. Kenneth Boulding in History and Theory 7 (1968): 83-90.

FALLACIES OF COMPOSITION

239

incomplete. The nation-state is still, in most chapters of Palmer's book,
the unit of analysis. He has advanced from national history to national
histories, but not to the next step, genuine supernational history.

Palmer's positive achievement, and Black's, should not be minim-
ized. They have both accomplished more in the way of diminishing
national stereotypes and national limits in historical scholarship than any
olher historians who come to mind. The limits of their accomplishment
are noted merely as a way of measuring the difficulty of the task.

Both Palmer and Black are Americans who specialize in aspects of
the history of Europe. Palmer is a historian of France; Black, of Rus-
sia. In the disparity between their own national origins and their na-
tional specialties both men gained a certain cosmopolitan advantage. The
worst cases of narrow nationalism in scholarship appear in the works of
scholars who, like myself, are American historians of America, or English
historians of England, or German historians of Germany. It is easy, in
these circumstances, to submerge all group types into the nation-state
without a second thought. There is no difficulty in finding an example —
indeed, it is difficult to find a survey history of the United States which is
not an example of cross-grouping in the form of submerging all groups
into the nation. Morison and Commager, Hofstadter, Miller and Aaron,
Williams, Current and Friedel — are all inveterate cross-grouping books.
And these are probably the best available surveys of American history.

If cross-grouping is most common in the form of confusing many
groups with nation-groups, it also appears in other forms. One of them
is the confusion of social and cultural groups of many kinds with geo-
graphical groups. Consider a hypothetical example: the inhabitants of
Manhattan, who are a geographical and a political group. But they are
not a society or a culture. Nevertheless it would be possible for an
anthropologist to attempt a study of the Manhattan Islanders, like the
Trobriand Islanders, and the result might have a certain superficial
plausibility. He could apply the rhetoric and the conceptualization of cul-
tural anthropology to the Manhattan Islanders, and illustrate it graphi-
cally and persuasively with many bits and scraps of "evidence." He could
begin by demonstrating that the Manhattan Islanders have an ecological
base and a political structure, complete with sachems and chieftains and
other picturesque functionaries. He could show that they have formed
traditional rivalries with other people — with the Staten Islanders, and
the Long Islanders, and the terrible Bronx People. He could "prove"
that they share certain physical characteristics in common — blackened
lungs, bloodshot eyes, bronchial congestion. He could locate a religious
common denominator in the fact that nearly all of them worship one God
at most (maximal monotheism, it might technically be called).

240

EXPLANATION

Our anthropologist concludes that children of the Manhattan
Islanders play curious ritualized games which are not widely known in
other "cultures" — stick ball, snatch-the-purse, buzz-the-fuzz, and other
jolly diversions. Adult Manhattan Islanders faithfully observe certain
strange taboos — custom forbids them to speak politely to a stranger, or
to smile in public, or to say please and thank you under any circum-
stances. They also keep curious dietary laws, somehow involving lox and
bagels. Their ethical practices are complex — stealing is forbidden to
private citizens, but encouraged in chiefs and sachems. Special burial
practices are customarily used for deviant Islanders, who are dropped
into the river after midnight, with their feet encased in concrete — a
probable symbol of disapproval.

All of this, of course, commits the fallacy of false culture. Man-
hattan is, I repeat, the name of an island, a borough, a geographical
group, a residential group, a political group, and maybe an ecological
group, but it is not the name of a cultural group. It would be a very great
mistake to confuse one with the other. And yet, as "culture" gradually
acquires the same mystique and the same elasticity among historians that
"nation" is slowly losing, the danger of the fallacy of false culture is
likely to grow.

Finally, we might note the existence of other, and special, forms
of cross-grouping, with reference to a single illustrative example. In the
introduction to this chapter we noted two different group types, among
others: voluntary associations and residence groups. There is a recent and
important book by Kenneth T. Jackson called Ku Klux Klan in the
City 1915-1930 (New York, 1967), which demonstrated what many
American historians had forgotten or never known — that the Klan was
an urban phenomenon as well as a rural one. But though Jackson's book
was a useful corrective to old errors, there was always a tension in his
conceptualization between the voluntary association called the Klan and
the residential group called the city. The Klan was predominant in
Jackson's inquiry, but most problems presented by the history of the
Klan are not susceptible to resolution within the city limits. Maybe there
is a cross-grouping fallacy here.

How are groups properly studied? In response to this question,
an immense methodological literature has been spawned by sociologists.
From the best of their work, historians have much to learn. No matter
what kind of group is studied, certain questions are likely to be relevant.
An able sociologist, Robert Merton, provides a useful checklist which
historians might employ. Merton suggests that the study of the structure

FALLACIES OF COMPOSITION

241

of groups involves the following problems:

1. Actual and expected duration of the group.

2. Actual and expected duration of membership within it.

3. Clarity or vagueness of definitions of membership.

4. Degree of engagement of members of the group.

5. Absolute size of the group and/or component parts.

6. Relative size of the group and/or component parts in reference
to other groups.

7. Open or closed character of the group.

8. Completeness of the group; i.e., the ratio of actual to potential
members.

9. Degree of differentiation; i.e., status and role as operationally
distinguished within the group.

10. Shape and height of stratification.

1 1 . Types and degrees of cohesion.

1 2. Potential for fission and unity within the group.

13. Extent and nature of interaction within the group.

14. Character of social relations obtaining in the group (see Par-
sons, The Social System, pp. 58-88).

15. Degree of conformity to group norms, toleration of deviant
behavior, and institutionalized departures from group norms.

1 6. System of normative controls.

17. Degree of visibility and observability within the group.

18. Ecological structure of the group.

19. Autonomy and dependence of the group.

20. Degree of group stability.

21. Modes of maintaining stability.

22. Relative social standing of groups.

23. Relative power of groups. 40

Merton is mostly interested here in group structure. Other questions
would have to be framed for group function and dysfunction, with respect
to members, nonmembers, and other groups. Moreover, a modification
is required to accommodate historical questions. All of Merton's ques-
tions can be set in motion and studied on a temporal axis. A major flaw
in sociology is its tendency to cast these problems in static forms. His-
torians, in this respect, can make a special contribution. But before they
can do so, they must make their answers to Merton's questions clear,
explicit, precise, and accurate.

Any student of groups must beware of many common kinds of

40. Merton, "Reference Groups and Social Structure," in Social Theory and Social
Structure, pp. 308-25.

242 EXPLANATION

error which lie along the way. He should be careful not to homogenize
individuals into groups, or one group into the group. Most people in
complex environments have belonged to many different groups — to con-
centric series of groups, and eccentric clusters of groups, and ephemeral
groups which appear and disappear within the span of a few moments. R.
M. Maclver has remarked that "All community is a matter of degree.
Our life falls within not one but many communities, and these stretch
around us grade by grade, building associations of every kind." The
homogenizing tendency has been encouraged by a good deal of romantic
humbug in radical and conservative thought, which persists in regarding
men as potential members of a single community. This ancient absurdity
is increasingly anachronistic. It is deleterious to empirical scholarship
and dangerous to peace and simple survival in the complex modern
world, where multiple group membership is a great irreversible fact of
life.

CHAPTER IX

FALLACIES OF
FALSE ANALOGY

The chief practical use of history is to deliver us from
plausible historical analogies.

— James Bryce

For epistemological puritans, analogies are not precisely explanations at
all. They are devices for discovering explanations. But given our loose
pragmatic everyday definition of explanation — i.e., "making clear, plain,
or understandable" — analogies are very useful explanatory tools. The
word "analogy," in modern usage, signifies an inference that if two or
more things agree in one respect, then they might also agree in another.
In its most elementary form, an analogy consists in a set of propositions
such as the following:

A resembles B in respect to the possession of the

property X.
A also possesses the property Y.
Therefore, it is inferred that B also possesses the

property Y.

The same thing can be said more succinctly in symbols:

AX : BX : : AY : BY

An unknown fourth term, BY, is thereby inferred from three known
terms, on the assumption that a symmetrical due ratio, or proportion,
exists.

Analogical inference plays an important, and even an indispensable,
part in the mysterious process of intellectual creativity. Many great in-
novating minds have, in the words of Jean Perrin, a French philosopher

244

EXPLANATION

of science, "possessed to an extraordinary degree, a sense of analogy." 1
The isochronous motion of a pendulum presented itself to Galileo in the
analogous behavior of a lamp swinging on its chain in the Pisa cathedral.
Recent scholarship has reinforced the legend of Sir Isaac Newton and the
great analogous apple. Benjamin Franklin operated by an analogy be-
tween electricity and a liquid; Huygens, by an analogy between ocean
waves, sound, and light; Van't Hoff, by an analogy between gases and
and solids in solution; Lord Kelvin, by an analogy between electricity
and heat; and Maxwell, by an analogy between light and electromagnet-
ism.

Analogies are equally useful and ornamental in the articulation of
ideas. They can do so in an internal way, by promoting an unconscious
or inchoate inference into the realm of rationality within a single mind. 2
And they also operate externally, as a vehicle for the transference of
thought from one mind to another. Analogies can brilliantly reinforce a
reasoned argument. They suggest and persuade, inform and illustrate,
communicate and clarify. They are versatile and effective pedagogical
tools. The great popularizers of science, from Voltaire to George Gamow,
could scarcely have operated without them.

Historians use analogies widely both as heuristic instruments for
empirical inquiry, as explanatory devices in their teaching, and as embel-
lishments in their writing. Often, analogies are used unconsciously — a
metaphor is an abridged form of analogy. Without analogies, creative
thought and communication as we know it would not be merely im-
practicable but inconceivable. The many uses of analogy, however, are
balanced by the mischief which arises from its abuse. Let us begin by
examining a few of them.

The fallacy of the insidious analogy is an unintended ana-
logical inference which is embedded in an author's language, and im-
planted in a reader's mind, by a subliminal process which is more power-
fully experienced than perceived. The mistake is a simple one, but serious
in its effects; for analogies are widespread in historical thought and im-
portant in the shaping of its content. Whenever a historian uses a
metaphor, he draws an analogy. And he uses metaphors all the time.
George Santayana perversely believed that all human discourse is meta-
phorical, which is surely an overstatement. But much more of our dis-
course is metaphorical than we are apt to realize. And the metaphors we

1. Quoted in Maurice Dorolle, Le Raisonnement par Analogie (Paris, 1949), p. 61.

2. John Williamson, "Realization and Unconscious Inference," Philosophy and Phenom-
enological Research 27 ( 1966) : 1 1-26.

FALLACIES OF FALSE ANALOGY

245

use to describe an object also determine the quality of our understanding
of it. Whenever an analogy is unconsciously used, so as to be dysfunc-
tional to that understanding, the fallacy of the insidious analogy results.

Historians instinctively employ many insidious analogies without
a second thought — or maybe even a first one. All of the following ex-
amples have caused trouble: Addled Parliament, Augustan age, avant-
garde, Axis, Babylonian captivity, Barnburners, blank check, Boxer,
Bloody Assizes, brinkmanship, Bubble Act, cameralism, capitalism,
Carbonari, Cold War, cordon sanitaire, Croix de Feu, Dark Ages, De-
pression, Digger, doughface, Enlightenment, Fabian, Fauve, Federalist,
feudalism, filibuster, Founding Father, Fronde, gag rule, gentlemen's
agreement, Good Neighbor Policy, Grand Peur, Guelph, Hats and Caps,
Heavenly Kingdom, imperialism, Industrial Revolution, Ironsides, Jac-
querie, jazz, jeremiad, Judas, Know-Nothing, Kulturkampf, Lebensraum,
Leveller, Loco-foco, logroller, Methodism, mother country, the Moun-
tain, muckraker, mugwump, New Light, Old Believer, Open Door,
papacy, Pact of Steel, puppet ruler, purge, Puritan, Quaker, quisling,
Reconstruction, Renaissance, revolution, Rump Parliament, Roi de
Soleil, Sea-Beggar, Spartacist, squatter, Take-Off, trust, Tory, the Sick
Man of Europe, underground, university, Utopia, vernacular, vigilante,
Village Hamden, wobbly, Whig, Xanthippe, yahoo, yellow-dog con-
tract, zambo, Zouave, Zionist.

Each of these terms contains within it an insidious analogy which
has served to distort our understanding of the object it is supposed
to describe. It would be absurd to suggest that any of these terms
should be stricken from the lexicon of history. They have been beaten
into our heads by many generations of well-meaning schoolmarms and
driven so deep they could not be removed even if we wished to do so.
One might, abstractly, wish to have a Jeffersonian revolution every
nineteen years in our historical vocabulary, to avoid becoming captives
of our language. But a more practicable solution would be for historians
themselves to search out the metaphors in their language and raise
them to the level of consciousness, where they can be controlled.

Other proper names are used in laymen's language as the first terms
in an analogical inference, with equally serious effects, of an opposite
nature. The common and customary meanings of Aristotelian, Benth-
amite, Ciceronian, Freudian, Jeffersonian, Machiavellian, Marxian, and
Platonic have diminished our understanding of the thought of these
men. Many a monograph on the Puritans has been motivated by a
determination to demonstrate that the common metaphorical meaning
of "puritanical" is seriously inaccurate as a description of the Puritans
proper. We are beginning to see a similar scholarly phenomenon with

246

EXPLANATION

respect to the term "Victorian." And yet, so powerful are these meta-
phors that even the monographs which seek to correct them become
captives, too, and commit the fallacy of the counterquestion by merely
reversing the objectionable implication.

There are still other insidious analogies in the verbs, adjectives,
adverbs, and even prepositions that historians conventionally use. Revo-
lutions tend to "break out," as if they were dangerous maniacs, locked
in a prison cell. Governments are overturned, like applecarts. Economies
boom and bust, like a cowboy on a Saturday spree. Cultures flower and
fade like a garden of forget-me-nots. Jefferson and Hamilton, or Pitt
and Fox, tend to "thrust and parry" through the history books, like
pairs of gentlemanly duelists. But Kennedy and Khrushchev, or Church-
ill and Hitler, bash and bludgeon like Friar Tuck and Little John.

Analogies of this sort are catching. And they serve to control
conceptualization. In histories of relations between Asia and the West,
door analogies are fashionable, as in Commodore Perry and the closed
door of Nippon, and the American Open Door Policy in China. In a
recent work on the history of China by an excellent Australian scholar,
one learns that "The Westerners banged heavily on the barred door of
the Chinese world; to the amazement of all, within and without, the
great structure, riddled by white ants, thereupon suddenly collapsed,
leaving the surprised Europeans still holding the door handle." 3 Such
analogies as this suggest that Asia is all structure and the West is all
function. They communicate a sense of clear and active purpose in the
latter and of mindless passivity in the former. Moreover, it is sometimes
assumed that China should swing freely before Western pressure, or
else it is slightly unhinged.

In the historiography of Poland, a different set of analogies is
customary. One is the traditional idea, deeply rooted in Polish literature,
that Poland is the "Christ among nations," a noble, transcendent being
which has suffered for the sins of all humanity, betrayed by the Jews
and crucified by the Romans. The result of this humbug is that history
becomes, in Namier's phrase, a visit of condolence. The Polish people
have been encouraged by their historians to develop a self-righteous
sense of persecution with few equals in the modern world. Every national
misfortune becomes a measure of the depravity of mankind — all man-
kind, that is, except the martyr nation, whose citizens are Poles apart.
This myth is profoundly dysfunctional to any constructive and statesman-
like attempt to deal with complex and critical diplomatic problems of
Eastern Europe.

3. C. P. Fitzgerald, The Birth of Communist China (Baltimore, 1964), p. 30.

FALLACIES OF FALSE ANALOGY

247

Other studies of Polish history tend to adopt a very different kind
of analogical imagery. It is historiographically conventional to compare
Poland to a bird — all feathers and fragile bones, big-beaked and small-
brained, beautiful but slightly weird, and sometimes a little sinister.
Stanley L. Sharp, a collector of many picturesque examples, declares
that "Ornithological comparisons seem traditional with reference to
Poland." He notes that

The ardent Polish nationalist Stanislaw Mackiewicz wrote in his critical study
of Beck's foreign policy, "Poles, like certain beautiful birds, are apt to lose
sight of their own surroundings, enraptured by their own song." . . . The
romantic poet Juliusz Slowacki once called Poland "the peacock and the
parrot of nations." The British writer John W. Wheeler-Bennett described
Poland's policy as that of "a canary who has persistently but unsuccessfully
endeavored to swallow two cats."

Sharp titled his own book, by the way, Poland, White Eagle on a Red
Field*

The complaint, in all of this, is not that analogies are used, but that
they are used insidiously, and that many absurd biases are bootlegged
into historical interpretations. An able scholar can, however, convert
an offense into an opportunity. He can study the analogies and meta-
phors which he instinctively invokes and thereby learn much about
the biases buried in his own mind, below the level of his consciousness.

We will never have historical writing without analogies. The next
generation of historians may perhaps learn to communicate with more
accuracy and precision by the use of mathematical symbols (unless
they are reduced by a nuclear catastrophe to a primitive exchange of
grunts and grimaces). But in either instance, there will still be analogies
and metaphors in historical discourse. Let us hope that they will be
developed with clarity, caution, and conscious reflection.

The fallacy of the perfect analogy consists in reasoning from
a partial resemblance between two entities to an entire and exact
correspondence. It is an erroneous inference from the fact that A and
B are similar in some respects to the false conclusion that they are the
same in all respects. One must always remember that an analogy, by its
very nature, is a similarity between two or more things which are in
other respects unlike. A "perfect analogy" is a contradiction in terms,
if perfection is understood, as it commonly is in this context, to imply
identity.

4. (Cambridge, 1953), p. 150.

248

EXPLANATION

This sort of error often appears in attempts at evaluation by
analogy, in arguments such as the following.

A and B are analogous in some respects.
A is generally a good thing.
Therefore, B is generally a good thing.

This set of propositions is structurally fallacious, for it shifts the analogy
from a partial resemblance to an identity, which is implied by the holistic
value judgment. If B were existentially analogous to A in respect to
X and Y, then it might be fairly though not conclusively inferred that
it is evaluatively analogous in the same limited sense. But it can never be
inferred that B is equivalent to A in either an existential or an evaluative
way.

Two examples of invalid historical analogies of this sort have
appeared in debates over American intervention in Vietnam. Spokes-
men for the United States government have tended to find an analogue
in Munich. A critic of the administration and its Vietnam policy, Arno
J. Mayer, has accurately criticized this unfortunate comparison, which
is, I think, not merely a rhetorical device, invoked by Washington
policy makers to justify their acts, but rather an operating assumption,
upon which their acts are based. Mayer protests that

By its proponents, the Munich analogy is designed to stress the identity, not
the similarity, of Hitler and Mao; of the Nazi German and the Communist
Chinese political systems and foreign policy objectives as well as methods;
and of externally incited subversion as well as the strategic significance of
Czechoslovakia and South Vietnam. The ensuing lesson is presented as self-
evident: no self-respecting American should want in the White House a
Chamberlain or Daladier, who by surrendering South Vietnam to the Chinese-
controlled North Vietnamese and Vietcong would encourage Peking to
activate its timetable for aggressive expansion into Southeast Asia and be-
yond. 5

Mayer proceeds to summarize the differences between Munich
and Vietnam: the disparity between the Vietcong and the Sudeten
Germans; the difference between the Czech government and the Saigon
regime; the difference between the strategic significance of Czechoslo-
vakia and Vietnam; the difference between the intentions of Nazi
Germany and Communist China; the difference between the military
capability of Anglo-French forces in 1938 and American power in the
late 1960s. Mayer also challenges the assumption that Hitler would
have changed his aggressive plans in any significant degree had the

5. Arno J. Mayer, "Vietnam Analogy: Greece, Not Munich," The Nation, March 25,
1968, pp. 407-10.

FALLACIES OF FALSE ANALOGY

249

allies stood their ground at Munich, and suggests that the only effective
deterrent would have been an effective alliance between Soviet Russia
and the Western nations, with rights of transit for Soviet troops through
Rumania and Poland. Such an alliance, he believes, was inconceivable,
given the intense and obsessive anti-Bolshevism of the Western powers.
Finally, Mayer denounces all "allegedly scholarly" historians and polit-
ical scientists who have "accepted, legitimized and propagated the cold
war eschatology according to which Nazism and Bolshevism were
essentially identical totalitarian systems bent on unlimited expansion by
a crude blend of outright force and externally engineered subversion."

Many details of Mayer's thesis are doubtful, as to his understanding
of both the Czechoslovakia crisis and the war in Vietnam. But his
protest is surely sound. There probably cannot be any sustained analogy
which will stretch from Munich to Saigon without breaking down.
But more important, there can never be an identical analogy, such
as Cold Warriors customarily draw between the 1930s and their own
predicament.

But Mayer is not done. He believes with E. H. Carr that the
"current era is exceptionally history-conscious" and that "today's citizen
has that pronounced need for and is peculiarly susceptible to analogies."
On this assumption, he concludes that a historian's duty consists not
merely in knocking over bad analogies but in setting up good ones, in
order to provide "the citizen with alternate historical sign posts." His
alternative to the Munich- Vietnam analogy is a Greece- Vietnam anal-
ogy, in which parallels are drawn between the "reticent role" of
Stalin and Mao; between indigenous Greek guerrillas and the Vietcong;
between Tito and Ho Chi Minh; between English retrenchment in
Greece and the French retreat from Vietnam; between the temporary
military and political weakness of Russia vis-a-vis the United States
in the late 1940s and the temporary weakness of China twenty years
later; between the domino theory of the Truman Doctrine and similar
assumptions in what might be called the Johnson Doctrine for Southeast
Asia. Mayer suggests that American policy — which includes contain-
ment of Communism, ordered modernization, and gradualist reform —
is similar in Greece and Vietnam. He implies that it has failed in Greece
and that it will fail in Southeast Asia as well. Moreover, "Not only
Greece — as the recent coup demonstrates — but also many of the develop-
ing countries lack the political integration, the social cohesion, and the
economic sinews to sustain gradual and ordered modernization and
reform, even with considerable foreign aid."

But Mayer has refuted one bad argument only to replace it with
a worse one. In his Greek analogue to Vietnam he commits the same

250

EXPLANATION

fallacy that others have done by analogizing from Munich to Southeast
Asia. Mayer concedes that there are "specific dissimilarities" between
Greece and Vietnam, but nowhere in his article does he specify them.
Instead, he tends to leap from analogy to identity, in the manner of his
opponents.

There are, of course, many major differences which he does not
take into account. Ho Chi Minh's concern with South Vietnam is of
a very different order from Tito's interest in Greece. The political
culture of Vietnam is far removed from that of Greece. The British
presence in Greece was of a different nature from the French regime
in Indo-China. American assistance to Greece was unlike our inter-
vention in Vietnam, both in quantity and in quality. Most important,
international political, military, and economic conditions have changed
radically from the late 1940s to the late 1960s. Vietnam is a painful
and difficult dilemma for the United States precisely because there is
nothing in our recent or distant past (or anybody else's) which is more
than incidentally and superficially similar.

Many analogues to Vietnam have been suggested — not merely
Munich and Greece, but the Mexican War, the Philippine Insurrection,
the Korean War, the insurgency in British Malaya, guerrilla warfare in
German-occupied Europe, the American Revolution, the Spanish rising
against Napoleon. In each of these instances, the analogy is very limited,
if indeed it exists at all. And there is surely no identity between any of
these happenings and the situation which American policy makers face
in Vietnam. That problem must be studied and solved in its own terms,
if it is to be solved at all. There are many particular historical lessons
which might be applied, in many limited and special ways, with due
allowance for intervening changes. There are restricted and controlled
analogies which might suggest hypothetical policy commitments for
possible use. But there are no comprehensive analogies which serve as a
short cut to a solution. A satisfactory historical approach to the problem
will not be oriented toward a search for an analogue but rather toward
a sense of environing continuities and changes within which the present
problem in Vietnam exists; combined with a keen and lively sense of
treacherous anachronisms and false analogies such as have deluded so
many well-meaning architects of American policy — and their critics, too.

There are many other examples of the identical analogy, a few of
which might be briefly noted. Ranke supported his government in the
Franco-Prussian war with the flat assertion that "We are fighting against
Louis XIV." This is a classic case of the abuse of historical knowledge.
A sophisticated sense of history consists not in the location of analogues

FALLACIES OF FALSE ANALOGY

251

such as this but rather in an ability to discriminate between sound
analogies and unsound ones.

Another quaint example, by an able historian who ought to have
known better, is the following assertion by Richard Pares: "It does
help us if we can realize that Charlemagne was just like an enlightened
American millionaire, for this recognition brings him into a class about
which we may know something." 6 This curious comparison may tell
us more about the extraordinary ideas which one British historian enter-
tained on the subject of enlightened American millionaires. And as it
stands, it is a false inference from resemblance to identity. Charlemagne
may or may not have been like an enlightened American millionaire in
some respect — though I cannot think of one, and Pares mentioned none
in particular. But he was surely not "just like" an American millionaire.
Therein lies a fallacy.

The fallacy of the false analogy is a structural form of error
which occurs when the analogical terms are shifted from one analogue
to another. Consider the following cases:

The second and third analogies are structurally sound. But the first
example is a false analogy in that there is an inconsistency between
X and Z.

This form of error is often exceedingly difficult to recognize,
because it is often hidden in semantical ambiguity, or buried in some
of the things which the author doesn't tell us. Let us consider an actual
example of this fallacy, perpetrated by Richard Morris. In an essay
called "Class Struggle and the American Revolution," 7 Morris addresses
himself to the sticky question of whether or not the War for Independence
was, by the design of its agents, a social revolution. He argues that it
was not directly, integrally, and aboriginally so, but rather engendered —
indirectly, incidentally, and gradually — a set of revolutionary social and
economic changes which were not among its "avowed objectives." This
argument is sustained by an analogy between the War of Independence
and the First World War.

1. AX : BZ :

2. AX : BX :

3. AZ : BZ :

: AY : BY
: AY : BY
: AY : BY

6. Richard Pares, The Historian's Business and Other Essays (Oxford, 1961), p. 8.

7. William and Mary Quarterly, 3d. ser. 19 (1962): 3-29.

252

EXPLANATION

An analogy might be fairly drawn to World War I [Morris writes]. Perhaps
the greatest change which came in the wake of that conflict, so far as America
was concerned, was the emancipation of American women, an extraordinary
phenomenon which liberated women from the home and thrust them into
the factory. The revolutionary impact of this social upheaval on postwar life,
politics, marriage, morals and the family is incalculable. And it never would
have happened so fast had it not been for the manpower shortage during
the war. But we have usually been taught that we went to war with Germany
over her renewal of unrestricted submarine warfare or because the House
of Morgan had floated loans to the Allies. I never realized that when
Woodrow Wilson called upon the Congress to declare war he really intended
to free American womanhood from the shackles of housework. Now within
certain limitations [unspecified by M.], I think the analogy to the American
Revolution is eminently fair. We did not declare our independence of
George III in order to reform the land laws, change the criminal codes,
spread popular education, or separate church and state. We broke with
England to achieve political independence, freedom from external controls,
emancipation, if you will, of the bourgeoisie from mercantile restraints. 8

Morris's analogy seems to reduce itself to the following four
propositions. The first three are factual. The fourth is an analogical
inference.

1 . World War I was a war which engendered revolutionary social
change in the United States.

2. The American War for Independence was a war which engen-
dered revolutionary social change in the United States.

3. Americans did not fight World War I to engender revolutionary
social change in the United States.

4. Therefore, it is inferred that Americans did not enter the War
for Independence to engender revolutionary social change in the United
States.

This looks structurally sound, on first inspection. But a closer look
suggests trouble. World War I was not the same kind of war as the
War of Independence — it was a total war, in which the nation was
enlisted with a degree of commitment which did probably not appear
in any eighteenth-century war, and certainly not in the American War for
Independence. And the engendering of revolutionary social change in
World War I is functionally connected to its total aspect. Moreover,
different processes of social change developed in the two cases. Morris's
first two propositions are disparate, in that they describe two different
things. They are to each other as AX is to BZ, rather than as AX is to
BX. Therein lies a fallacy.

8. P. 26.

FALLACIES OF FALSE ANALOGY

253

The fallacy of the absurd analogy is another structural form
of analogical error, in which an inference is extended between two
nonrelated characteristics. Consider two hypothetical examples:

This rubber ball and that apple are both

red, round, smooth, and shiny.
That apple is very good to eat.
Therefore, this rubber ball will be very good to eat.

Secondly:

This rubber ball and that apple are both red, round,

smooth, and shiny.
That apple looks pretty in a Christmas stocking.
Therefore, this rubber ball will look pretty in a Christmas

stocking.

The first of these analogies is patently absurd. But the second, given
certain aesthetic assumptions, is correct. The difference between them
is that the qualities of the ball and the apple described in the first terms
of the analogy are functionally relevant to aesthetics but not to edibility.
There is, in short, a rule of relevance in analogizing, which must always
be respected. In our elementary form:

AX : BX : : AY : BY

There must be a relationship between X and Y if there can be an analogy
between A and B."

The English historian G. M. Trevelyan recalls in his autobiography
a character named Edward Bowen, an "eccentric genius" of "somewhat
ascetic habits" who was Trevelyan's housemaster at school. But Bowen's
genius did not consist in a talent for analogical inference. Trevelyan
remembered that "He once said to me, some years after I had left school,
'O boy, you oughtn't to have a hot bath twice a week; you'll get like the
later Romans, boy.' " 10

The fallacy of the multiple analogy is a structural deficiency
which occurs when a second analogy is bootlegged into the main analogy
so as to undercut the basis of comparison. Consider the following
hypothetical example, which comes from the work of an English phi-

9. For a suggestive discussion, see C. Mason Myers, 'The Circular Use of Metaphor,"
Philosophy and Phenomenological Research 26 (1965-66): 391-402.

10. G. M. Trevelyan, An Autobiography and Other Essays (London, 1949), p. 11.

254

EXPLANATION

losopher, Alfred Sidgwick: "The growing size of London bodes evil to
England because London is the heart of England and a swollen heart
is a sign of disease." 11

This statement might be broken down into three parts:

1. London is analogous to a heart (presumably in the sense that
both perform a vital circulatory function).

2. A swollen heart is a sign of disease.

3. The growth of London bodes evil to England.

But between the second and the third statements, two other analogies
are tacitly added:

2.1 Swelling is analogous to growing.

2.2 A sign of disease is analogous to that which bodes evil for
England.

Assuming that an analogy is merely a partial resemblance and not an
identity, neither of these two tacit pairs of analogues is interchangeable.
There is, therefore, no continuity from proposition two to proposition
three. The trouble is papered over by semantical ambiguity in the
original statement, an ambiguity which serves to camouflage the addi-
tional analogies.

A historical example appears in George Rude's The Crowd in
History, in which the author solemnly asserts that "Thus, beheaded, the
sans culotte movement died a sudden death; and having, like the cactus,
burst into full bloom at the very point of its extinction, it never rose
again." 12 This statement combines three disparate analogies. It is
objectionable on both stylistic and substantive grounds. As a mixed
metaphor, it is a literary monstrosity. As a multiple analogy, it is a
logical absurdity. Many amusing examples appear from time to time in
The New Yorker. The major complaint to be entered against these
excrescences is not aesthetic but analytical. Vulgarity can coexist with
empiricism; illogic cannot.

The fallacy of the holistic analogy is, I think, the fatal fallacy
of metahistory, as it has been practiced by Spengler and Toynbee and
a host of others. It is an attempt to construct an analogical inference
from some part of history — to the whole of history. All metahistorians
have built their interpretations upon a metaphor, for there is nothing
else at hand. Empiricism is impossible if the object is to tell the whole

U. Alfred Sidgwick, Fallacies (London, 1883), p. 179.
12. P 106.

FALLACIES OF FALSE ANALOGY

255

truth. Only some nonempirical method of inference, such as analogy,
can be used.

A close student of analogy, Harald Hoffding, has observed that

if analogy is employed metaphysically or cosmologically, it is not a single
realm of Being serving to illuminate another single realm; it is a single realm
that is used to express Being as a totality. This symbolism is of a different
kind and has a different validity from that brought to bear on particular
fields. It cannot be carried out to its full consequences and it cannot be
verified. ... In these respects, cosmological and metaphysical symbols are
different from scientific ones. . . . Religious symbols share the fate of the
metaphysical. In both cases the attempt is made to create absolutely valid
final concepts; the only difference lies in the motive. 13

The behavior of analogy in cosmology, metaphysics, and religion
is the same as its behavior in metahistory. But in the latter, claims to
empirical accuracy are entered. Empiricism fails, however, in the face
of holistic problems, and the analogy alone is left to carry the weight.
Arnold Toynbee has been fairly and fully criticized by many reviewers
for this mythological use of analogy in A Study of History. He has
entered a plea of guilty, but only to certain "excesses." The criticism,
however, cuts deeper than that: it alleges that Toynbee's method is
fundamentally analogical, and his analogies are fundamentally unsound,
because they cannot be put to the test. To this, of course, Toynbee does
not plead guilty, for he cannot, without repudiating the work of a life-
time."

The fallacy of proof by analogy is a functional form of error,
which violates a cardinal rule of analogical inference — analogy is a
useful tool of historical understanding only as an auxiliary to proof. It
is never a substitute for it, however great the temptation may be or how-
ever difficult the empirical task at hand may seem.

Humanity appears to have made a little progress in this respect. A
student of Renaissance culture has written, "While modern thought is
fully aware of the tentative nature of analogical reasoning, earlier
thought tended to consider an analogy as an end in itself and to rest
content in an aesthetic and essentially poetic awareness of the feeling
of understanding the analogy brought." 15

13. Harald Hoffding, The Problems of Philosophy (New York, 1913), p. 121; and
Der Begriff der Analogic (Leipzig, 1914), passim.

14. Arnold J. Toynbee, A Study of History, vol. 12, Reconsiderations (New York,
1961), pp. 30-41.

15. Joseph A. Mazzeo, "Analogy and Renaissance Culture," Journal of the History
of Ideas 15 (1954): 299-304. See also, Thomas De Vio, Cardinal Cajetan, The Analogy

256

EXPLANATION

But the progress is incomplete. So successful are analogies in
creating the illusion of sense and certainty that they are widely used as
a method of proof in their own right. I have heard a sociologist argue
that, though an analogy never affords a "rigorous demonstration," it
may nevertheless provide an "appreciable coefficient of affirmation,"
which can be cast in terms of probability. This is solemn nonsense. Ana-
logical probability is altogether as elusive as analogical certainty, in the
absence of an empirical test. The accuracy of that empirical test may
be cast in probabilistic terms with precision, but not the analogy itself,
which has finished its work after the empirical level is reached.

An example of this fallacy, in which an analogy is not transcended,
is a controversial essay on slavery and Negro personality by Stanley
Elkins — a work of which we have taken note several times. 16 Elkins
establishes an analogy between two different institutions — plantation
slavery in Anglo-America and concentration camps in Nazi Germany.
The latter have been studied by many psychologists who were interested
in the personality patterns the camps caused in their inhabitants. Elkins
argues that the camps and slavery were analogous in several respects and
that slavery created a "Sambo" personality which is comparable to the
"old prisoner" mentality which some psychologists have found in the
concentration camps.

Elkins's argument is plausible and highly persuasive. His analogy
operates effectively as a heuristic device in his own inquiry and as a
rhetorical instrument in his presentation. It suggests much but — it proves
nothing. One might argue that his analogy is structurally imperfect in
a variety of ways, and that the institutional parallels between slavery and
concentration camps tend to dissolve on close inspection. But there is a
more serious complaint to be made against Elkins's work. He does not
move beyond his analogical insight to establish empirically the existence
of the Sambo personality pattern. There are only a few casual snippets
of impressionistic evidence, much of which is secondary or tertiary.
Elkins has insisted that he did not mean to prove his argument by
analogy, but he nevertheless does so implicitly in his book.

In my opinion, there is an important truth in Elkins's thesis. Many
other historians seem to think so, too. The argument, analogy and all,
is beginning to work its way into the textbooks, and even into historical
novels, such as William Styron's The Confessions of Nat Turner, which
appears to owe a special debt to Stanley Elkins and which may serve

of Names (1498), trans. E. A. Bushinski (Pittsburgh, 1953), a systematization of the
so-called Thomistic theory of analogy.

16. Stanley Elkins, Slavery: A Problem in American Institutional and Intellectual
Life, rev. ed. (New York, 1963).

FALLACIES OF FALSE ANALOGY

257

to popularize his thesis. One of Elkins's students has even ground out a
monograph, which echoes the master's expectations in the spirit of Sambo
himself. But everything still hangs precariously upon an analogy, which,
even if it were the best analogy in the world, would be insufficient to
sustain it.

The fallacy of prediction by analogy occurs when analogy
is used to anticipate future events — as it often is, in the absence of
anything better. H. W. Fowler observed that analogy "is perhaps the
basis of most human conclusions, its liability to error being compensated
for by the frequency with which it is the only form of reasoning avail-
able." 17

The trouble with futurist analogies is not that they might be wrong,
but rather that they must be utterly untestable and inconclusive. The
problem is not that there is a probability of error within them, but that
there is an indeterminancy of probability. It is not possible to distinguish
a true historical analogy from a false one without an empirical test of
its inference. As long as one of those parts remains in the future, the
analogy is untestable.

A historiographical case in point is a collection of quasi-historical
essays edited by Bruce Mazlish and published as The Railroad and the
Space Program: An Exploration in Historical Analogy (Cambridge,
Mass., 1965). Mazlish and his colleagues seriously attempted to estimate
the future effect of the space program upon American society by means
of an analogy with the past effects of the railroad in nineteenth-century
America. The contributors were able scholars all, and their essays
uniformly reached a high level of sustained and sophisticated cerebration.
But with respect to the future consequences of the space program, they
might as well have hired a gypsy to study the palm of Werner von Braun
or invited an astrologer to contribute a paper to their project. Their
conclusions about the space program are either tenuous in the extreme,
or truistic, or else Delphic utterances of the sort which confidently
predict with considerable semantical confusion that maybe X will
happen, or maybe it won't.

The work of Mazlish and his colleagues, in short, is not very useful
for serious students of the space program. But, significantly, the book is
highly suggestive for students of the railroads. Most contributors devote
much of their interest to the latter. The hypothetical heuristic construct
provided by the space program has a stimulative effect in historical

17. H. W. Fowler, Modern English Usage (Oxford, 1926), p. 20.

258

EXPLANATION

inquiry, which is altogether independent of its truth value with respect
to the space program itself. It provides many suggestive hints and
hypotheses which might be put to an empirical test by an economic
historian, with the possibility of new and important insights into
economic development in the nineteenth century. In short, the Mazlish
volume demonstrates explicitly a truth long implicit in the operations of
historians — namely, that an analogy is a useful device for a sort of retro-
diction of past events and for the generation of hypothetical interpreta-
tions which can be put to the test. One can reason from an idea of the
future (however mistaken it may prove to be) to an insight into the
past, and put the latter to the test. But the process is not reversible.

Mazlish might reply that there are no empirical ways of knowing
the future. But this, I think, is a mistake. Two other methods are em-
ployed with increasing accuracy in a wide range of fields — in meteor-
ology, economics, and demography. These methods are both historical
in nature. One of them consists in the discovery of past trends and their
extrapolation into the future, in some cases with determinable degrees
of probability. The other is a kind of theoretical knowledge, or con-
ditional knowledge, which takes the form of "If, then" propositions —
empirical propositions which are tested by reference to past events.
Forecasting of this sort can work — indeed, it does work — even with
respect to events which are partly determined by willful acts of reasoning
agents.

But a prediction by analogy is useless in itself. Sometimes the
analogizer covers himself in the fashion of Mark Twain's weather fore-
caster: "Probable nor'east to sou'west winds, varying to the southard
and westard and eastard and points between; high and low barometer,
sweeping round from place to place, probable areas of rain, snow, hail,
and drought, succeeded or preceded by earthquakes with thunder and
lightning." 18

Nothing else can improve his accuracy.

The misuses of analogy are many and complex, but all fallacies
in this chapter can be divided into two groups. First, there are structural
fallacies of analogical inference — analogies which are imperfect in
their form. Second, there are functional fallacies, in which sound
analogies are applied to inappropriate purposes.

Any intelligent use of analogy must begin with a sense of its limits.
An analogical inference between A and B presumes that those two

18. Quoted in D. S. Halacy, The Weather Changers (New York, 1968), p. 30.

FALLACIES OF FALSE ANALOGY

259

objects are similar in some respects but dissimilar in others. If there
were no dissimilarities, we would have an identity rather than an
analogy. Analogical inference alone is powerless to resolve the critical
problem of whether any particular point is a point of similarity or
dissimilarity. It can never prove that because A and B are alike in respect
to X, they are therefore alike in respect to Y. Proof requires either
inductive evidence that Y exists in both cases, or else a sound deductive
argument for the coexistence of X and Y. If either of these attempts at
proof is successful, then the argument becomes more than merely analog-
ical. If neither is successful, there is no argument at all.

In common practice, some deductive inference as to the connection
between X and Y is commonly drawn. In empirical inquiry, an attempt
must also be made to establish the existence of X and Y. Galileo, in the
example of the analogy between the chain lamp and the motion of a
pendulum, mentioned at the beginning of this chapter, immediately
advanced beyond analogy to empiricism, by means of an experiment
which he cleverly contrived on the spur of the moment. He timed the
swings of the cathedral lamp by his own pulse beat. In the same fashion,
Newton and Franklin and the others quickly proceeded to put their
analogies to the test.

The psychological power of analogical explanation is dangerous
both to logic and to empiricism. Many bad ideas have had a long life
because of a good (effective) analogy. If analogy is used to persuade
without proof, or to indoctrinate without understanding, or to settle an
empirical question without empirical evidence, then it is misused. Some-
times the results are not merely disagreeable but downright dangerous.
In the formation of postnuclear public policy, nothing is quite as lethal
as a faulty prenuclear analogy. Fallacies of this sort are apt to be failures
not of will but of understanding. In public questions of nuclear policy,
they may be the last thing a well-meaning statesman ever intends to
commit — the very last thing.

PART III

ARGUMENT

CHAPTER X

FALLACIES OF
SEMANTICAL
DISTORTION

Rightly to be great is not to stir without great argument.
— Shakespeare

The fallacies in this chapter are not particular to professional historians.
But every historian runs the risk of committing them. All historical inter-
pretations are arguments, and they must conform to a logic of argumenta-
tion if they are to cohere as truth. Many so-called historical facts are
arguments, too. When a scholar makes a "factual mistake," he may
actually have committed what is commonly called a fallacy of argument.

By argument, of course, I do not mean a controversy or a dis-
agreement, but rather an attempt to proceed from premises to a con-
clusion by orderly and rational inference. Anyone who hopes to
accomplish this object can get himself into three different kinds of
trouble — structural distortion, semantical deception, and substantive
distraction.

The first of these categories refers to formal fallacies of argument.
There are many different systems of formal logic in the world — the
canonical logic of ancient China, the grammatical logic of ancient India,
the syllogistic logic of Aristotle, the dichotomous logic of Ramus, and
the mathematical logic of Russell and Whitehead are among the most
familiar. Surely there are more logical systems in heaven and earth than
are dreamt of by any logician. If a visitor from another planet were to
descend upon our world, he might bring with him a formal logic unlike

264

ARGUMENT

anything an earthling knows. It is probable — indeed, almost certain —
that our posterity will produce formal logical systems far different from
those we now possess.

This book contains no discussion of formal logic for two reasons.
First, every existing logical system has been fully discussed by logicians
with more expertise than a historian can hope to attain. Second, and
more significant, no presently articulated system of formal logic is really
very relevant to the work historians do. The probable explanation is
not that historical thought is nonlogical or illogical or sublogical or anti-
logical, but rather, I think, that it conforms in a tacit way to a formal
logic which good historians sense but cannot see. Some day somebody
will discover it, and when that happens, history and formal logic will be
reconciled by a process of mutual refinement. But at present, we must
confine ourselves to a more humble task.

This chapter is about informal fallacies of a semantical nature.
Many a ponderous volume has been produced on the subject of seman-
tics. Our business, however, is with a small part of this large problem —
the attainment of clarity and precision in historical prose. Imprecision in
a sonnet may be the soul of art. But in a closely argued scholarly state-
ment, its consequences are always inconvenient, and sometimes fatal. The
first question to be asked of any expression, in every historical context, is
not "Is it true?" but "Is it meaningful?" A meaningful expression, for
present purposes, is merely one which successfully serves two purposes.
First, it clearly and consistently communicates an author's intention to
a reader's understanding. Second, it communicates, in the same fashion,
an author's intention to his own understanding. The latter object is often
more elusive than the former. The semantical deceptions in this chapter
are mostly self-deceptions rather than deliberate attempts to mislead.

Meaning, in this limited sense, is mainly a definitional problem,
which presents no difficulty in most ordinary words and word combina-
tions. It is rarely necessary for a historian to analyze a statement such
as "What is meaning?" in the manner of G. E. Moore, and ask, "What
is the meaning of meaning' 1 . What is the meaning of what! What, indeed,
is the meaning of is?" Most words in a historiographical argument are
sufficiently clarified in a tacit way by the argument itself. In a famous
phrase by P. W. Bridgman, "The true meaning of a term is to be found
by observing what a man does with it, not what he says about it." 1

Usually, what a man does with a term is a sufficient guide. Nothing
more needs to be said. But some people do extraordinary things with
their terms, and a good deal of confusion can result if definitions are not

1. Percy W. Bridgman, The Logic of Modern Physics (New York, 1946), p.7.

FALLACIES OF SEMANTICAL DISTORTION 265

explicitly made and consistently used. We shall begin, in this chapter,
with a review of the common fallacies of semantical deception, and then
proceed to consider the process of definition itself in a more constructive
way.

The fallacy of ambiguity consists in the use of a word or an
expression which has two or more possible meanings, without sufficient
specification of which meaning is intended. In politics and poker,
ambiguity has its uses, and even its justifications. Adlai Stevenson, in
the presidential campaign of 1 952, told a tale about his namesake, who
was a candidate for Vice President in 1892:

I have read a lot of stories about the time when my grandfather campaigned
in the state of Washington for the Vice Presidency, exactly sixty years ago
this month. The big issue, I am told, at that time was whether your majestic
mountain was to be named Mount Tacoma or Mount Rainier. Apparently
that was the only subject of interest in Washington at that time. Anyway, the
views of Seattle and Tacoma were in violent disagreement and it seems that
my adroit grandfather solved this difficulty by giving each audience from the
back platform of his train an eloquent speech about the beauties of the
mountain, and then went on to say, "And I want everyone to know, all of
you good people, that I emphatically agree that this magnificent mountain
should be named — " and just then they pulled the whistle on the train and
it started with a puff, and the old man bowed to the audience graciously and
they cheered ecstatically. 2

But when the proverbial whistle is pulled in a train of scholarly argu-
ment, the results are not amusing. Simple semantical ambiguity develops
in several different ways. It sometimes arises from the use of historical
terms which have meant so many different things that they are meaning-
less until a specific meaning is clarified in context. A sophisticated stylist
is able to accomplish this end tacitly, without a laborious and pedantic
definition of terms. But an explicit formal definition, however clumsy it
may be, is always preferable in historical writing to the most graceful
ambiguity. 3 Words such as democracy, capitalism, nationalism, class,
culture, education, party, feudalism, and romanticism — to name but a
few — should never be employed without an ad hoc definition. An elab-
orate exercise in philology is rarely required — merely a simple working
definition, or perhaps a set of definitions, which will provide a rapport
between an author and his readers.

Ambiguity also creates confusion in the relationship between a

2. Adlai Stevenson, Major Campaign Speeches (New York, 1953), p. 83.

3. Ambiguity can be a thing of beauty, in belletristic prose. See William Empson,
Seven Types of Ambiguity, 3d ed. (Norfolk, Conn., 1953), for an extended discussion.

266

ARGUMENT

researcher and his sources. Giles Constable got into trouble in a mono-
graph on monastic tithes in the early Middle Ages. A reviewer of his
work complained that "More emphasis might also have been placed upon
the basic ambiguity that persistently hovers around the word 'tithe,'
particularly in private acts. The term could also mean a secular quitrent,
and in scores if not hundreds of private documents, the historian simply
cannot tell what type of rent the term is signifying." 4

It is easy to shrug off such objections as this, as mere nitpicking
pedantry. But the results can be destructive to otherwise excellent scholar-
ship.

Another kind of ambiguity consists in the use of an old term in a
new way without warning. Sir Lewis Namier provides an amusing
example of a Victorian lady who complained that she did not like a
house because it was "very romantic." Her correspondent responded,
"I don't understand why you should wish it not to be very romantic."
The Victorian lady replied, "When I said romantic I meant damp." 5

There is much of this sort of ambiguity in the writings of the so-
called New Left in America. Every red-blooded radical knows that
American libertarianism is really totalitarianism, and that American
democracy is really tyranny, and that American freedom is really a
species of slavery, and that American tolerance is really intolerance, and
that an open society in the United States is really a gigantic concentra-
tion camp with invisible barbed wire. To the uninitiated, the meanings of
these words appear to be bent beyond the breaking point. Perhaps the
confusion is owing to the stubborn determination of American radicals
to conceptualize their own society in terms taken from a very different
one. The irrelevance of these terms is hidden by means of ambiguity
from the people who invoke them.

A special form of ambiguity might be called the fallacy of etceter-
ation. It occurs when a historian, enumerating evidence, or categories,
or types, or reasons, seeks to disguise a problem, or perhaps the poverty
of his materials, by employing the abbreviation "etc." An example is
supplied by Maitland, who discovered that Elizabeth I was the first
English monarch to add the words et cetera to her title. Maitland pon-
dered the problem, and finally came to the following conclusion:

No doubt she is Defender of the Faith, though we cannot be sure what
faith she will defend. But is that all? Is she or is she not Supreme Head upon

4. David Herlihy in The American Historical Review 71 (1965): 134, reviewing Giles
Constable, Monastic Tithes: From Their Origins to the Twelfth Century (New York,
1964).

5. Sir Lewis Namier, "History and Political Culture," in Fritz Stern, ed., Varieties
of History (New York, 1956), p. 386.

FALLACIES OF SEMANTICAL DISTORTION

267

earth of the Church of England and Ireland? ... It was a difficult problem.
On both sides there were men with extreme opinions, who, however, agreed
in holding that the solution was not to be found in any earthly statute book.
. . . Then a happy thought occurs. Let her highness etceterate herself. This
will leave her hands free, and then afterwards she can explain the etceteration
as occasion may require. And so, Queen Elizabeth I became the first English
ruler to be "solemnly etceterated."

The "et cetera" probably meant, "and (if future events shall so decide,
but not further or otherwise) of the Church of England and also of
Ireland upon earth the Supreme Head." 6

Other Englishmen solemnly etceterated George Washington during
the American Revolution. In July, 1776, the Howe brothers addressed
a peace feeler to "George Washington, Esq., etc., etc." The intended
recipient refused to accept it, on the ground that he was General
Washington. The senders blandly replied that "etc., etc." included all
titles which Squire George chose to adopt. The fighting continued, in
a decidedly unambiguous way. 7

The fallacy of amphiboly arises in an argument where mean-
ing is muddled by slovenly syntax — bad grammar, or poor punctuation,
or both. Trouble of this sort commonly develops in three specific ways.
First, it may derive from a relative pronoun with more than one possible
reference. The classic example is "He said, 'Saddle me the ass.' And they
saddled him." Second, amphiboly may arise from a misplaced modifier,
as in the proverbial definition of anthropology as "the science of man
embracing woman." Third, amphiboly may be the result of an elliptical
construction, as in a wartime poster which urged everyone to "Save Soap
and Waste Paper." The following historiographical examples are hypo-
thetical:

1 . "Richly carved Chippendale furniture was produced by colonial
craftsmen with curved legs and claw feet."

2. "Many Americans were outraged when President Theodore
Roosevelt had a Negro for dinner."

3. "The measures of the New Deal were understandably popular,
for many men received jobs, and women also."

4. "The ship was christened by Mrs. Coolidge. The lines of her
bottom were admired by an enthusiastic crowd."

Sometimes an amphibolous operation is performed on a primary
document by a careless editor. He makes a small change in grammar

6. Frederic W. Maitland, "Elizabethan Gleanings," Collected Papers, 3:157-65.

7. John R. Alden, A History of the American Revolution (New York, 1969), p. 265.

268

ARGUMENT

or punctuation, with a major alteration in meaning. Two nineteenth-
century historians, Henry Cabot Lodge and Henry Adams, both inaccu-
rately quoted a letter from a New England Federalist, Stephen Higginson,
to Timothy Pickering, on the subject of disunion, as follows: "I have
seen your letters to Mr. Cabot and Mr. Lyman on the question of sepa-
ration, which is a very delicate and important one, considered in the
abstract. We all agree there can be no doubt of its being desirable." But
the manuscript actually read, "I have seen your letters to Mr. Cabot and
Mr. Lyman on the question of separation, which is a very delicate and
important one. Considered in the abstract we all agree there can be no
doubt of its being desirable." Two petty changes in punctuation were
made, together with other errors in other parts of the letter, with the
result of making Higginson appear more favorable to disunion than in
fact he was. 8

The fallacy of figures is a form of ambiguity which consists
in the abuse of figurative language, so that a reader cannot tell whether
or not a literal meaning is intended; or if so, what that meaning might
be. There are many figures of speech which are commonly used in
conventional discourse without a second thought. Henry Peacham, in
The Garden of Eloquence (London, 1577), named and defined more
than two hundred of them, most of which are still often used, but
not always recognized by the user. In another edition of the same work,
Peacham provided "cautions" against the abuse of each figure — abuse
which has become more common than correct and effective employment
of these rhetorical devices.

Today, Henry Peacham's subject seems altogether as quaint as
many of his examples. It has, indeed, been increasingly unfashionable
since the late sixteenth century. Nevertheless, it is one which a historian
might study carefully, for two purposes. First he can improve his
style by controlled and conscious use of these embellishments. There is
nothing necessarily fussy or false or weak about them. The greatest
classics of Elizabethan and Jacobean literature, such as the King James
Bible and the plays of Shakespeare, gained refinement, power, and hon-
esty from frequent use of figures.

Second, a study of figures should help a historian to keep them
out of his prose when he wishes to do so. Sometimes we must make

8. The letter, dated March 17, 1804, appears in Lodge, The Life and Letters of George
Cabot (Boston, 1877), p. 453; and Henry B. Adams, ed., Documents Relating to New
England Federalism, 1800-1850 (New York, 1965), pp. 361-62. The original is in
the Pickering Papers, Massachusetts Historical Society.

FALLACIES OF SEMANTICAL DISTORTION 269

our writing as literal as possible. Clarity often requires the ruthless
removal of every stylistic ornament from our prose. And yet these
embellishments are so deeply rooted in our language — owing partly to
the genius of Elizabethan literature, and partly to the nature of language
itself — that it is as difficult to pull them out as it is to put them in.
The following list (much shorter than Peacham's) includes a few figures
which can cause great confusion in historical writing. Definitions and
examples are taken from Fowler's Modern English Usage, from Thrall
and Hibbard's Handbook to Literature, and from Henry Peacham.

Antiphrasis, the use of a word to convey its opposite meaning, as
Antony's "such honourable men."

Apophasis, the making of an assertion while seeming to deny it.
In the American Historical Review, 71 (1965): 147, one historian
wrote of another, "It is no real criticism to say that the author does not
equal Eileen Power in her ability to make the past come alive."

Aposiopesis, a sudden breaking off, in which the reader is required
to supply the missing words, as "If he should fail — "

Catachresis, the misapplication of a word, perhaps for effect or
possibly from ignorance, as "chronic" for "serious," or "decimate" for
"slaughter," or "dilemma" for "difficulty," or "dock" for "pier," or
"ratification" for the Senate's consent to a treaty.

Ellipsis, the omission of words necessary to the grammatical struc-
ture of a sentence, as Pope's "Where wigs [strive] with wigs, [where]
with sword-knots sword-knots strive,/ [Where] Beaus banish Beaus, and
[where] coaches coaches drive."

Euphemism, a word in which accuracy is sacrificed to taste, as
"remains" for "corpse," or "passed on" for "died."

Hypallage, an inverted relationship between several words, as
Virgil's "the trumpet's Tuscan blare," for "the Tuscan trumpet's blare."

Hyperbole, an exaggeration for effect, as "older than time."

Hysteron Proteron, the reversal of the logical order of ideas, as
Dogberry's, "Masters, it is proved already that you are little better than
false knaves, and it will go near to be thought so shortly."

Litotes, a double negative, or multiple negative, as in a sentence
by Harold Laski: "I am not, indeed, sure whether it is not true to say
that the Milton who once seemed not unlike a seventeenth-century Shel-
ley had not become, out of an experience ever more bitter each year,
more alien to the founder of that Jesuit sect which nothing could induce
him to tolerate." This statement is discussed in a splendid essay by
George Orwell, "Politics and the English Language," in A Collection
of Essays (New York, 1954), pp. 162-76.

Meiosis, an understatement, as "altercation" for "fight."

270

ARGUMENT

Metonymy, tilt, use of a substitute term for another which is sug-
gested by it, as "The Pentagon announced today . . ."

Oxymoron, the joining of contradictory terms, as "cruelly kind," or
"simply gorgeous," or "frightfully nice," or "make haste slowly," or
"ordered confusion," or "deafening silence," or Michael Walzer's char-
acterization of John Calvin's thought as "a theology antitheological,"
in The Revolution of the Saints (Cambridge, 1965), p. 24.

Persiflage, defined by H. W. Fowler as "irresponsible talk, of
which the hearer is to make what he can without the right to suppose
that the speaker means what he seems to say." An example is the argu-
ment of Carl Becker in The Heavenly City of the Eighteenth Century
Philosophers (New Haven, 1932), pp. 30-31. The book is disguised
as a historical interpretation, and yet on the flyleaf of a presentation
copy, he wrote, "This certainly isn't history. I hope it's philosophy, be-
cause if it's not it's probably moonshine: — or would you say the
distinction is oversubtle?"

Prolepsis, describing an event as happening before it could have
done so. Fowler provides a poetic example: "So the two brothers and
their murder'd man/ Rode past fair Florence."

Prosopopoeia, by which a person is made to represent a type, or
an idea, or an institution, or a quality, as in "Hamiltonian" for Federal-
ist, or "a Jefferson" for any democratic thinker.

Suggestio Falsi, a statement which is literally true, but which
encourages a false inference, as to the question, "Is X a competent
scholar?" the answer, "X is my friend; I'd rather not say."

Solecism, a violation of grammar or idiom, as "Winston tastes
good, like a cigarette should." Richard Sherry, in A Treatise of Schemes
and Tropes (1550, Gainesville, Fla., 1961), p. 36, defined it as "an
unmete and unconvenient joynynge together of the partes of speech . . .
which because it is used by famous authors, instead of faults, be
called figures."

Soroesmus, in Peacham's phrase, "a mingling together of divers
languages, as when there is in one sentence English, Latinne & French."

Synecdoche, a figure in which a part is used for the whole, as
"bread" for "food" or "man" for "people."

Zeugma, the use of a single modifier for two terms, with one of
which it seems logically connected: "See Pan with flocks, with fruits
Pomona crowned."

If all this seems a little remote from the daily business of a historian,
consider the following common uses of hyperbole, in which the author is
often fooled as well as his readers. Historians have been known to write
"always" for "sometimes," and "sometimes" for "occasionally," and

FALLACIES OF SEMANTICAL DISTORTION

271

"occasionally" for "rarely," and "rarely" for "once." In historical writing
"certainly" sometimes means "probably," and "probably" means "pos-
sibly," and "possibly" means "conceivably."

Similarly the phrase "It needs no comment" should sometimes
be translated as "I do not know what comment it needs." When a his-
torian writes, "It is unknown," he might mean "It is unknown to me,"
or "I don't know," or even "I won't tell." The expression "in fact" some-
times means merely "in my opinion." And the phrases "doubtless" or
"undoubtedly," or "beyond the shadow of a doubt" sometimes really
should be read, "An element of doubt exists which I, the author, shall
disregard."

Another familiar variation on this melancholy theme is the ten-
dency to convert the verdict of a historian into "the verdict of history."
Arthur Schlesinger, Jr., seems to be habitually attached to this ugly
usage. 9 And when he writes, in the same volume, "And so Landon passed
into history" (p. 232) he means, "I'm finished with Alf for a while."

Another figure of the similar sort is the hyperbole "every schoolboy
knows." If any schoolboy knew all the things which Macaulay believed
that every schoolboy knows, then that omniscient child might be awarded
a university degree, honoris causa. "Every schoolboy knows . . . who
strangled Atahualpa," Macaulay wrote. I doubt that any modern school-
boy outside of Peru even knows who Atahualpa was, much less how
he died. Maybe things were different when Prescott was popular, and
there were not so many strangulations to remember. But generally
speaking, the phrase "every schoolboy knows" means "some very
learned scholars have forgotten, or failed to emphasize sufficiently, the
fact that."

All these figurative expressions are difficult to interpret in practice,
because they are sometimes used figuratively and sometimes literally,
and often there is no signal to tell the reader which sort of meaning is in-
tended. It would, of course, be absurd to remove all figures of speech from
historical usage merely because some figurative language is apt to
confuse more than it clarifies. Figures are so deeply embedded in our
language that they cannot be removed, even if one wished to do so.
And they are not merely ornamental but also useful in enlarging the
range of communication. But wherever they may be taken literally by an
intelligent reader, and a literal reading of them falsifies or distorts the
meaning, then the fallacy of figures results.

The fallacy of accent occurs in an argument when meaning
is distorted by emphasis. Imagine a tabloid headline reading:
9. See, e.g. The Coming of the New Deal (Boston, 1959), p. 175, passim.

272

ARGUMENT

CATASTROPHIC CRIME WAVE IN CITY

predicted by prelate

Another example is provided by a historian of Salem, Massachusetts,
who writes that

Captain L had a first mate who was at times addicted to the use of

strong drink, and occasionally, as the slang has it, "got full." The ship was
lying in a port in China, and the mate had been on shore and had there in-
dulged rather freely in some of the vile compounds common in Chinese
ports. He came on board, "drunk as a lord," and thought he had a mortgage
on the whole world. The captain, who rarely ever touched liquors himself,
was greatly disturbed by the disgraceful conduct of his officer, particularly
as the crew had all observed his condition. One of the duties of the first
officer [i.e., the mate] is to write up the "log" each day, but as that worthy
was not able to do it, the captain made the proper entry, but added: "The
mate was drunk all day." The ship left port the next day and the mate got
"sobered off." He attended to his writing at the proper time, but was ap-
palled when he saw what the captain had done. He went on deck, and soon
after the following colloquy took place:

"Cap'n, why did you write in the log yesterday that I was drunk all day?"

"It was true, wasn't it?"

"Yes, but what will the owners say if they see it? T will hurt me with
them."

But the mate could get nothing more from the captain than "It was
true, wasn't it?

The next day, when the captain was examining the book, he found at
the bottom of the mate's entry of observation, course, winds, and tides:
"The captain was sober all day." 10

Adlai Stevenson liked to tell a tale upon himself, which involves the
fallacy of accent. During the 1956 presidential campaign he arrived at
a Chicago airport to find a shouting mob waiting for him. In the front
rank was an immensely pregnant lady, carrying a large sign reading

"STEVENSON IS THE MAN." 11

A historiographical example is suggested in an article by M. Rostov-
tzeff, "The Decay of the Ancient World and Its Economic Explanations."
The author examined the accuracy of interpretations of "the decline of
ancient civilization," and sensibly concluded that much depends upon the
way in which that phrase is accented. To speak of the decline of ancient
civilization is one thing. But to talk of the decline of ancient civilization
is quite another. "Thus," Rostovtzeff wrote, "to apply to events in the
ancient world in the centuries after Diocletian and Constantine the term

10. Charles E. Trow, The Old Shipmasters of Salem (New York, 1905), pp. 14-15.

11. Kenneth S. Davis, The Politics of Honor: A Biography of Adlai Stevenson (New
York, 1967), p. 363.

FALLACIES OF SEMANTICAL DISTORTION

273

'decay' or 'decline' is unfair and misleading. If, however, in the formula
'decay of ancient civilization' we lay stress on 'ancient' and not on 'civil-
ization,' the formula hits the mark." 12

Another example, this time from American history, has developed
from a famous phrase in the Declaration of Independence, "All men
are created equal." One wonders if Jefferson meant "All men are
created equal" or "All men are created equal." If the former, then he
might have agreed with Nathaniel Ames, who observed that

All men are created equal,
But differ greatly in the sequel.

If he intended the latter, however, a more extended idea of equality
appears.

On the afternoon of July 4, 1826, the fiftieth anniversary of the
Declaration of Independence, Jefferson and John Adams dramatically
died within a few hours of each other. Adams's last words were spoken
at noon, when he awakened briefly and said, "Thomas Jefferson sur-
vives," and slipped back into unconsciousness. Some historians have
made this utterance into a noble tribute to the third President. But one
wonders what precisely Adams intended by his delphic phrase. Maybe
he said it with a groan, and meant, "Is that sandy-haired son of a bitch
still alive?"

Innuendo is a form of the fallacy of accent. There are many
refined examples of insinuated reflections on character and reputation
in the writings of Tacitus. One student of his work has commented
on "his readiness to tax to the uttermost every resource of Latin in
the cause of antithesis or innuendo." 13 A less subtle specimen, which
derives a certain illustrative clarity from its coarseness, was committed
by a nineteenth-century New England clergyman and was described
in a recent biography of the Grimke sisters:

Traveling on a boat to New Haven, [James G.l Birney met the Reverend
Leonard Bacon, pastor of the First Congregational Church of New Haven,
and took occasion to question him about a remark attributed to him regard-
ing Sarah Grimke.

"And what was that?" asked Rev. Bacon.

"I have been told," said Birney, "that in speaking of fanaticism, at one
time in New England, you said a Quaker woman had been known publicly to
walk through the streets of Salem, naked as she was born — But that Miss
Grimke had not been known to make such an exhibition of herself yet. D'J
you say this?"

12. M. Rostovtzeff, "The Decay of the Ancient World and Its Economic Explanations,"
Economic History Review 2 (1930): 197-99.

13. John Jackson, Introduction to Tacitus, The Annals (London, 1931 ), p. 238.

274

ARGUMENT

"I did," Bacon readily admitted and added after a pause: "And should
I have said that she did?" 14

If innuendo is understood in a more general way to mean any
insinuation or connotation, then it cannot be eliminated from language.
Some have tried to do so, by desperate expedients. It is said that Jeremy
Bentham was so fearful of the fallacy of accent that he deliberately
employed a reader with a perfectly monotonous voice. But even a mono-
tone can commit a fallacy of accent in the rendering of a style which
requires tonal variations, as all English style must. The only rational
expedient is to strive for controlled connotation — even, one might say,
controlled innuendo — so that veiled meanings are properly aligned with
the author's purpose, and with reason and empirical accuracy. The
fallacy of accent consists not in the use of emphasis, but in its unfair
and inaccurate use.

$»» The fallacy of equivocation occurs whenever a term is used in
two or more senses within a single argument, so that a conclusion ap-
pears to follow when in fact it does not. Sometimes, equivocation is
deliberately employed for stylistic effect, without a significant substantive
distortion. As such, it is the basis of much humor of a low kind. Witness
the following "bon mot" by a "certain witty Lord" in eighteenth-century
London: "His Lordship being informed, that a Lady, lately divorced,

would probably be married to the Earl of Upper O- y, said that it

is about time she was upper O y for she has been under O y long

enough." 18

Other equivocations are not intended. In a recent study in Sudanese
history, Byron Farwell attempted to explain the alleged corruption of the
Mahdi by the following extraordinary enthymeme: "Power, of course,
corrupts, and luxury when available is a powerful temptation." 1 ' There-
fore, by Farwell's logic, when luxury became available to the Mahdi, he
became powerfully corrupt.

A more serious modern instance appears in Herbert Aptheker's
American Negro Slave Revolts (2d ed. [New York, 1963]). Aptheker
meant to prove that there were many revolts in the history of American
Negro slavery — a revolt being defined at the outset as something involv-
ing "a minimum of ten slaves" with "freedom as its object." He concluded
that there were some two hundred fifty of these happenings in less than

14. Gerda Lerner, The Grimke Sisters from South Carolina (Boston, 1967), p. 196.

15. D. Judson Milburn, The Age of Wit (New York, 1966), p. 45.

16. Byron Farwell, Prisoners of the Mahdi (New York, 1967), p. 24.

FALLACIES OF SEMANTICAL DISTORTION

275

two hundred years. But if his original definition of revolts is respected,
the number shrinks to fifteen or twenty — for as Aptheker proceeds, he
loosens his definition of slave revolts to include events involving fewer
than ten people, risings not directed toward freedom, revolts in French
and Spanish colonies, conspiracies, and alleged conspiracies. Had Ap-
theker's definition been rigorously applied, his thesis would have been
visibly untenable. Instead, it is obscurely so.

$»» The fallacy of quibbling is a form of equivocation which in-
volves two or more people in a single argumentative exchange. It occurs
whenever the meaning of a term is changed as it changes hands, with a
resultant argumentative distortion.

An intricate example appears in an answer to Max Weber's thesis in
The Protestant Ethic and the Spirit of Capitalism, by a Swedish economic
historian named Kurt Samuelsson. Weber's argument for a functional
relationship between capitalism and the Protestant ethic rested upon a
careful definition of these terms. Samuelsson subtly shifted those defini-
tions in his attempt at refutation. In the end, he succeeded in refuting
a thesis, but not the Weber thesis.

Weber defined capitalism in a special and limited sense, to mean
the rational organization of formally free labor. 17 He specifically re-
pudiated the more general and more common definition of capitalism as
"the impulse to acquisition, pursuit of gain, of money, of the greatest
possible amount of money." But Samuelsson substituted this latter defini-
tion for Weber's in his version of the Weber thesis, defining capitalism as
"large scale accumulation," or "the growth of capital mobilisation and
credit provision on a large scale." 18

Samuelsson had no difficulty in demonstrating that there were
many capitalists, in his sense, who had nothing to do with Protestantism.
But this fact did not contradict Max Weber's argument — indeed, Weber
specifically affirmed it. Samuelsson's work is useful in one respect, at
least. Many others have also misread Weber; specifically, they have mis-
understood it as an argument for a connection between Protestantism
and the heaping up of riches, or Protestantism and the growth of the
acquisitive impulse, or Protestantism and the growth of capital mobili-
zation. To misread Weber in this way is to miss his point by a country

17. Max Weber, The Protestant Ethic and the Spirit of Capitalism (New York, 1948),
p. 21.

18. Kurt Samuelsson, Religion and Economic Action, trans E. Geoffrey French
(New York, 1961), pp. 84, 92.

276

ARGUMENT

mile. Samuelsson's iconoclastic essay might usefully serve as a reminder
of this fact, and as a remedy for a vulgar version of the Weber thesis which
has circulated widely in the universities.

$*> The black-or-white fallacy is a form of error which occurs in
the misconstruction of vague terms — i.e., terms such as hot and cold,
light and dark, good and bad, free and unfree, and right-wing and left-
wing. There is no firm and fixed criterion for distinguishing between
hot and cold, no sharp line which separates these two words, but an
area of doubt between them. A precise distinction can be made for a
particular purpose, but only by drawing an arbitrary line. And that
practice is dangerous in two different ways.

First, it is sometimes argued that because such a distinction is
arbitrary, and merely a matter of degree, that no "real" distinction exists.
In other words, it might be asserted that there is no "real" difference
between two different shades of gray, because they are both called gray,
and only an arbitrary line can separate light gray from dark gray, the
only "real" difference being between white and black. It is sometimes
said, for example, that there is no "real" difference between the treatment
of Negroes in Mississippi and Massachusetts, because Mississippi's record
in this respect is not pitch-black, and Massachusetts' is far from being
snow-white. But nevertheless, there is a difference between two shades
of gray, a difference which this line of argument disguises by a semantical
trick.

Conversely, it is possible to err in the opposite way and be deluded
by language into a reification of the arbitrary line, in which case dark
gray becomes black, and light gray becomes white. An example was
recently perpetrated by two able American historians, Lee Benson and
James P. Shenton. Both these scholars were enthusiastic supporters of
Senator Eugene McCarthy and opponents of Senator Robert F. Kennedy.
On March 20, 1968, the two historians published in The New York
Times an advertisement which read as follows:

The Responsibility of American Intellectuals is to tell the truth. Always. The
truth is: The movement that has made Senator McCarthy its symbol ex-
emplifies rationality, courage, morality. The movement Senator Kennedy
commands exemplifies irrationality, opportunism, amorality. The truth is:
To be moral and remain moral, movements must always choose moral men
and moral means. The truth is: The end never justifies the means. Never. If
American intellectuals do not know that, they have learned nothing from
history. The truth is: In March 1968, history has caught up with American
intellectuals. They must choose between morality and amorality, between
McCarthy or Kennedy. And to act on their choice. Publicly. Unequivocally.
Immediately.

FALLACIES OF SEMANTICAL DISTORTION 277

The confusion in this extraordinary statement consists not merely in the
fact that politicians do not behave according to these Manichaean ex-
pectations, but that language doesn't behave this way either. "McCarthy"
and "Kennedy" are words of one sort; "moral" and "amoral" are of
quite another. The latter are vague words, like hot and cold, which
are qualities which exist in various degrees, and in various respects. If
an arbitrary distinction is drawn between "moral" and "amoral," then
it must be drawn clearly and explicitly, and its arbitrariness must be
borne in mind. If it is not, two shades of gray are converted by semantical
mumbo jumbo into black and white. When words are used as they are
by Benson and Shenton, they become meaningless. Their statement is
not merely false — it is solemn and literal nonsense.

Another example of this form of the black-or-white fallacy occurs
in the preface to Sir Lewis Namier's The Structure of Politics at the Ac-
cession of George 111 (2d ed. [London, 1957]), in which the author de-
clared, "I refrain from adding a bibliography. There can be none for
the life of a community; I hardly remember having come across con-
temporary materials, or any book reproducing such materials, which did
not contribute something to my information." 19 In other words, "I can-
not list all of my sources; therefore, I need not list any of my sources."
Everyone who has done the sort of research which Namier did will credit
his statement that every piece of English evidence which was contempo-
rary to his English subject became a source for his study. But some
sources were more productive than others. A select bibliography of the
most useful material would have been a good, gray middle ground be-
tween the black of no bibliography, and the white of a complete one. 20

The remedy for all these informal fallacies is formal defini-
tion faithfully applied. There are many different kinds of formal defini-
tion. A historian ought to choose consciously and carefully from the
range of possibilities available to him. Imprecision results not merely
from an incomplete or inaccurate or inconsistent definition, but also
from the use of an inappropriate definitional type. Suppose, for example,
a definition of "Quaker" is required. It could be one or more of the fol-
lowing types, some of which overlap:

1. A definition by genus and difference locates a term within a
larger class, and then supplies specific differences. Alan Simpson employs

19. P. xiv.

20. For a similar error, see William Appleman Williams's justification of the omission
of footnotes from The Contours of American History (Cleveland, 1961), p. 491.

278

ARGUMENT

this method when he defines a Quaker as a member of the left wing of
Puritanism. (Puritanism in Old and New England, p. 1.)

2. A theoretical definition might include a statement of principles
involved in an idea. Frederick Tolles defines a Quaker as a man whose
religious faith "represented an equilibrium of four elements — mysticism,
prophetism, perfectionism, and universalism." (Meeting House and
Counting House, p. 6.)

3. A lexical definition defines a word by explaining its common,
ordinary, or accepted usage. In Webster, Quaker is defined as, "in com-
mon usage, a member of the Society of Friends." In seventeenth-century
Massachusetts, however, Quaker was also used as a synonym for fanatic.

4. A stipulative definition introduces a wholly new expression into
the language, or gives a new and special meaning to an old expression,
as in the use of Quaker to mean "a dummy gun," or "a guessing game
played with coins." (Mitford Matthews, Dictionary of Americanisms,
p. 1337.)

5. A precising definition defines a word in a specific way for a
special purpose. Imagine a handyman who is instructed to remove
all the brush from a back lot, but to leave all the trees. In his instructions,
a tree might be defined as anything more than three inches thick at the
base. Similarly, a historian of the Pennsylvania Assembly in the mid-
eighteenth century could conceivably define a Quaker as "an assembly-
man who belonged to a political faction, of which the acknowledged
head was Isaac Norris II."

6. An enumerative definition provides a complete list of every item
to which a word applies. It is impracticable, of course, in the case of
"Quaker," where it would have to be "George Fox, William Penn," etc.,
until all Quakers are named. For other purposes, however, such as de-
fining the Big Three, an enumerative definition is natural and normal.

7. An ostensive definition is more generally useful, for it is a repre-
sentative rather than a complete listing. Thus, a Quaker is "such a person
as George Fox, or William Penn, or John Woolman."

8. A genetic definition describes the origin of the thing designated
by a word: a Quaker is "a member of a sect founded (1647) by the
English religious leader George Fox. . . ." (Thomas H. Johnson, The
Oxford Companion to American History, p. 315.)

9 . A constructive definition tells how a thing can be made — which
is not very appropriate to the Quaker example, though it is often used in
other contexts. But it might be applied even to "Quaker," as "a person
who is admitted to a Society of Friends, either by means of birth to a
member, or by signifying his willingness to join."

10. An operational definition specifies the tests which determine

FALLACIES OF SEMANTICAL DISTORTION

279

whether or not a term applies to the thing in question. For example, "a
person may be called a Quaker if he manifests the following behavioral
characteristics: a tendency to gather in silent meetings, to dress plainly,
to address others as thee, to refuse to doff his hat except in prayer, to
refuse to bear arms, and to refuse to swear oaths."

11. A synonymous definition explains the meaning of a term by
identifying other terms with the same meaning; as, "Quaker: Friend,
Foxite, etc."

12. An analytical definition defines a thing by detailing its parts.
"A Quaker believes in the doctrine of the Inward Light, the separation
of church and state, the irrelevance of priests and creeds, the right and
duty of conscientious objection, the responsibility of philanthropy and
charity," etc.

13. A synthetic definition defines a thing by reference to other
things. Robert Doherty used this method in a recent monograph, The
Hicksite Separation (New Brunswick, 1967), in which he compiled
such charts as the following, for Quakers and Non-Quakers in Philadel-
phia, circa 1828:

wealth: real estate

Assessed value Quaker Non-Quaker

of real estate

Number

Percent

Number

Percent

42

30

214

46

$ 1-999

18

13

92

20

1000-1999

16

11

46

10

2000-2999

11

8

41

9

3000-3999

10

7

29

6

4000-

43

30

43

9

Total

140

99

465

100

14. A persuasive definition defines a term in such a way as to induce
a person to accept or reject some principle or value. An example would
be John Morley's definition of Quakerism as "the most devout of all
endeavours to turn Christianity into the religion of Christ." {Oliver
Cromwell, p. 429.) Another example, on the other side, is the traditional
definition of a Quaker as "a man who prays for his neighbors one day
a week, and preys on them the other six."

15. A figurative definition defines a term in metaphorical terms,
as, William Penn's definition of Quakers as "dissenters in our own land,"
or "the children of light," etc.

The term Quaker is one which presents comparatively few defini-
tional difficulties. The reader is invited to try to define "Puritan," on the

280

ARGUMENT

one hand, or "Hicksite," on the other. Historical definitions can be trou-
blesome not merely because of the complexity of the historical process,
but also because of the complexity of the definitional process. There is as
much confusion between structural types of definition as between sub-
stantive definitions themselves.

A historian, however, will simplify his definitional responsibilities
by observing a few simple but often neglected rules. First, he does not
have to define a term for all eternity but merely for the limited context
and purpose he has in mind. Failure to respect this common-sense prin-
ciple has occasioned a good deal of semantical imprecision, and even some
imprecision in the first degree — which is to say, willful and premeditated
imprecision. A Dutch historian, Jan Romein, has written,

The historian always works with vague notions. Race, people, nation, state.
Nobility, bourgeoisie, small-middle class, proletariat. Republic, monarchy,
dictatorship, democracy. Feudalism, capitalism, socialism, fascism. Renais-
sance, baroque, romanticism, liberalism. Where is the historian whose hand
will not hesitate when venturing to define any of these conceptions? But
where, also, is the historian who will refrain from using them as being too
vague? The historian must work with vague notions because his object does
not admit of exact ones. 21

Romein used this argument to justify his favorite imprecision, the
idea of a Zeitgeist, or spirit of an age. But his argument is absurd. The
defining hand of the historian might well hesitate if indeed he had to
define Romein's illustrative terms for all seasons, independently of his-
torical context and historiographical purpose. But he is merely required
to establish workable ad hoc definitions, which can be done with maximal
precision and minimal effort.

Second, a historian should, where possible, define historical pro-
cesses rather than nonhistorical states — things that happen rather than
things that are. Many historians have labored long and hard to estab-
lish a definition of a static thing called political democracy, an atemporal
entity which involves them not merely in vast metaphysical confusion
but in a good deal of historical bewilderment as well — for political
democracy has meant different things at different times. Their task is
considerably eased if they think in terms of process, and conceptualize
in terms of "democratization" rather than "democracy." I have found
that historians will agree quickly on a workable definition of political
democratization, with reference to a particular project — as, for example,
the expansion of voter participation in an increasingly free electoral
process; but they will argue for years over the meaning of "political
democracy."

21. Quoted in Pieter Geyl, Encounters in History (New York, 1961), p. 326.

FALLACIES OF SEMANTICAL DISTORTION 281

Third, contrary to the opinions of many logicians, a historian needs
rarely to concern himself with finding a definition which is entirely com-
mensurate or equivalent to the thing denned. And he should never try
to define the "essential" qualities of a thing. The first is supremely
difficult; the second is a superstition. Let us recur to the example of the
handyman in the back lot, who was instructed to cut down all the
brush but not to destroy any trees. For that purpose, a definition of a
tree as anything more than three inches thick at the base is sufficiently
precise, and perfectly satisfactory — though in many another context, to
define a tree as anything more than three inches thick at the base is not
merely false but meaningless.

Fourth, historians should supply an enumerative or ostensive defi-
nition for every doubtful proper noun of prominent importance in their
writing. Often other kinds of definition are necessary, too. But these, I
think, are always required, for a historian must establish clear links
between language and reality, between the conceptual and the concrete.
Consider, for example, the case of Herbert Butterfield's The Whig In-
terpretation of History. Butterfield's analytical terms are very clear, and
for a historian, extraordinarily exact. But E. H. Carr has complained of
the book that "though it denounced the Whig interpretation over some
130 pages, it did not, so far as I can discover without the help of an
index, name a single Whig except Fox, who was no historian, or a single
historian save Acton, who was no Whig." 22 Carr is not quite correct.
Hallam is mentioned on page 4, and Hallam was a historian and a
Whig, though he was both just barely. Two other names appear on
page 96 of Butterfield's book. But these exceptions neither vitiate Carr's
criticism, nor mitigate Butterfield's offense.

Fifth, though a historian is not required, as has been suggested, to
find a fixed definition of a term for all time, he must find a definition
which will remain constant and consistent through a single argumenta-
tive series. If he shifts to another argument in a single work, then the
term might be changed in that transition. But it cannot be changed within
an argument, if the term is functional to the reasoning process.

Sixth, a historian must openly and directly confront the problem of
connotation and innuendo. Neither can be altogether eliminated. Words
will always communicate meanings on many different levels. A historian
should turn connotation and innuendo to constructive purposes. He
should seek to control and direct them, rather than to remove them — to
align the many-leveled meanings of his words with each other, and with
his evidence, and with his logic.

22. Carr, What Is History?, p. 50.

CHAPTER XI

FALLACIES OF
SUBSTANTIVE
DISTRACTION

Nay, Sir, argument is argument. You cannot help paying
regard to their arguments if they are good. If it were testi-
mony, you might disregard it. . . . Testimony is like an
arrow shot from a long bow; the force of it depends on the
strength of the hand that draws it. Argument is like an arrow
shot from a cross bow, which has equal force though shot
by a child.

— Samuel Johnson

A second set of informal fallacies are substantive rather than semantical.
They all operate by shifting attention from a reasoned argument to other
things which are irrelevant and often irrational. A definitive catalogue
of distractions is, of course, impossible and absurd. The following sixteen
fallacies are merely a few common forms of error. Nearly all of them
have been discussed before. The standard Latin nomenclature is adopted,
for the sake of convenience.

The two most full and ample discussions in print are a collection of
posthumous essays by William G. "Single-Speech" Hamilton called
Parliamentary Logic (1808) and Jeremy Bentham's Book of Fallacies
(1824). 1 Bentham's work is perhaps the more important of the two,

1. Both works have been reprinted, and Bentham's has appeared in different versions
under different titles. The best is Bentham's Handbook of Political Fallacies, ed. Harold
A. Larrabee (Baltimore, 1932). There are many other works of the same sort.
Etienne Dumont's Traite des Sophistries politiques (Geneva, 1816) was a redaction
prepared from Bentham's own unpublished notes. More recently Arthur Schopenhauer
composed a similar work called The Art of Controversy which is available in an
English translation. Of many modern works, Robert H. Thouless, Straight and Crooked
Thinking (New York, 1932) and L. Susan Stebbing, Thinking to Some Purpose (Balti-
more, 1938), might be specially recommended.

FALLACIES OF SUBSTANTIVE DISTRACTION 283

though Hamilton's is more readable, and in some technical respects more
sound.

There are two serious errors in Bentham's work. First, he tended to
assume that these forms of error are usually evidence of some sort of
sinister interest in their authors. "Is it credible . . . that their inanity
and absurdity should not be fully manifest to the persons who employ
them?" he asked. "No," he answered in his solemn way, "it is not
credible." 2 But this is a very great mistake. Many of the following exam-
ples, if not quite all of them, are clearly not the result of a deliberate
attempt to deceive but rather of obscured understanding by authors
who were themselves deceived — a condition which is far more common
than Bentham was prepared to admit.

Bentham's second mistake was equally serious. He tended to assume
that these fallacies were such that "their application affords a presumption
either of weakness or of total lack of relevant arguments on the side
on which they are employed." 3 But many a bad argument has been used
in a good cause. It would be a very profound and pedantical mistake
to presume that any of the fallacies in this book, if they appear in a
historical interpretation, are prima-facie evidence that the interpreta-
tion is false in all respects, and utterly useless. 4

The fallacy of argument ad verecundiam is an appeal to
authority. The conventional Latin label means literally an argument to
modesty, or shyness, or shame. This form of error is an egregious but
effective rhetorical technique which puts an opponent in the awkward
position of appearing to commit the sin of pride if he persists in his op-
position.

The most crude and ugly form of an argument ad verecundiam in
historical writing is an appeal to professional status. David Donald, for
example, published an essay in which he attempted to analyze the
leadership of the abolitionist movement in social terms, and concluded
that most abolitionists were "descended from old and socially dominant
Northeastern families, reared in the faith of aggressive piety and moral
endeavor ... an elite without function, a displaced class in American
society . . . basically abolitionism should be considered the anguished

2. Bentham, Handbook of Political Fallacies, ed. Larrabee, p. 242.

3. Ibid., p. 227.

4. Bentham's book is mistaken in many other ways, which are less relevant to our
inquiry. He believed that Parliamentary government was so constituted as to encourage
the commission of fallacies, and that the congressional system of the American
Republic was much superior in this respect. To this infatuation, a student of American
politics can only add a sigh.

284

ARGUMENT

protest of an aggrieved class against a world they never made." 5

Shordy after this thesis appeared in print, a young graduate student,
Robert Skotheim, published a reply 6 in which he criticized Donald's
sample of abolitionist leaders, the imprecision of his data, the absence
of a control, and a false extrapolation from abolitionist leadership to all
abolitionists.

Donald, unfortunately, lost his temper. His rebuttal consisted of a
series of ignoratii, punctuated by an ad verecundiam. And he finished
with the following salvo, aimed at Skotheim in particular, and uppity
graduate students in general.

Mr. Skotheim's criticisms raise the general problem, which Professor David
M. Potter so cogently discussed not long ago in a review in the Journal of
Southern History, of "how far it is possible to proceed in judging historians'
interpretations without grounding these judgments in an understanding of the
history which is being interpreted." Perhaps the teaching profession is at
fault in encouraging young scholars like Mr. Skotheim to undertake studies
in methodology and historiography before he has demonstrated his com-
petence in research. 7

Donald's unfortunate argument from professional status calls to
mind an exchange in Lewis Carroll's Sylvie and Bruno:

"Do you mean to say," said Lady Muriel, "that these manikins of an
inch high are to argue with me?"

"Surely. Surely!" said the Earl. "An argument doesn't depend for its
logical force on the size of the creature that utters it!"

She tossed her head indignantly. "I would not argue with any man less
than six inches high!" she cried. "I'd make him workl"

"What at?" said Arthur, listening to all this nonsense with an amused
smile.

"Embroidery!" she readily replied. "What lovely embroidery they
would do!" 8

In historiography, such crude forms of argument ad verecundiam are
rarely to be met with — in print, at least. The explanation is not that
scholars are gentlemen, but rather, as Bolingbroke noted many years
ago, that "those who are not such, however, have taken care to appear
such in their writings." 9

5. David Donald, Lincoln Reconsidered (New York, 1956), pp. 19-36: "Toward a
Reconsideration of Abolitionists."

6. Journal of Southern History 25 (1959): 356-65.

7. Ibid., 26 (1960): 156-57.

8. The Complete Works of Lewis Carroll, Modem Library ed. (New York, n.d.), p. 450.

9. Henry Bolingbroke, Letters on the Study and the Use of History (London, 1870),
pp. 10 f.

FALLACIES OF SUBSTANTIVE DISTRACTION

285

$»> More common and more subtle forms of argument ad vere-
cundiam appear in appeals to all the paraphernalia of pedantry. Among
them are:

1 . Appeals to pedantic words and phrases

2. Appeals to references

3. Appeals to quotations

4. Appeals to length

5. Appeals to detail and specificity

6. Appeals to mathematical symbols

The first of these forms of error is committed by scholars who never
use a little word when a big one will do. Historians take a certain pride
in their alleged immunity from this fallacy — in their freedom from jargon
and academic affectation. But their conceit is not correct; indeed, it is
growing increasingly inaccurate as an understanding of contemporary
historiographical language. Ordinary everyday words like "simple" are
replaced by monstrosities such as "simplistic" without any refinement of
meaning. Special fields of historical inquiry are building pedantic vo-
cabularies at an appalling rate. Urban historians, for instance, speak
endlessly of "urbitecture," "areal differentiation," "ecosystems," "nodal
points," "metropolitan matrices," "ruralization," "subareal mosaics,"
"conurbation," and other such neologisms, which are in some cases
useful for their precision and defensible for their utility. But these terms
are also used for purposes of legitimization, as ritual incantations which
serve to camouflage doubt, confusion, illogic, imprecision, and igno-
rance. One recent book about urban history blithely defines several of
these terms in several different ways, and adds an unblushing explanation
that "there is little consistency in the use of metropolitan nomenclature,
and we have not attempted to be unduly precise here." 10 If professional
jargon is imprecise, then I think it is utterly indefensible. There can be
no complaint, of course, against the use of technical terms, but merely
against their illegitimate use for effect.

Sometimes I suspect that Americans are particularly susceptible to
this unfortunate fallacy. Bertrand Russell, in his memoirs, has recalled:

In Chicago I had a large seminar, where I continued to lecture on the same
subject as at Oxford, namely, "Words and Facts." But I was told that Ameri-
cans would not respect my lectures if I used monosyllables, so I altered the
title to something like "The Correlation between Oral and Somatic Motor
Habits." Under this title, or something of the sort, the seminar was ap-
proved."

10. Charles N. Glaab and A. Theodore Brown, A History of Urban America (New
York, 1967), p. 272n.

11. Bertrand Russell, Autobiography, 1914-1944 (Boston, 1968), p. 331.

286

ARGUMENT

Be that as it may, there are many examples of labored pedantry which
are equally objectionable for reasons of logic, rhetoric, and empiricism.
One reads such a sentence as the following: "The philosophical anthro-
pology pointed to here can perhaps best be described as a phenomeno-
logical anthropology: that is to say, a logos of the phenomenon of
anthropos, which has its source and its 'subject matter' in the concernful
questing for the being of man-in-quest himself," 12 and one wonders why.

This humbug is common enough in history today. It will probably
become more so. Already, we find a simple mutiny, on Drake's first cir-
cumnavigation voyage, ponderously and perhaps falsely discussed as a
manifestation of status jealousy. 13 The pioneer of many fashionable his-
toriographical techniques today might be heard in this connection. "The
ABC of my trade," Marc Bloch wrote, "consists in avoiding big-sounding
abstract terms. Those who teach history should be continually concerned
with the task of seeking the solid and concrete behind the empty and the
abstract.""

A second variety of an argument of the authority of learning symbols
is named the "fallacy of references" by an irate philosopher 13 who was
properly infuriated by the regrettable tendency of pedants of all persua-
sions to lard a lean thesis with fat footnotes which are irrelevant, or super-
fluous, or even something unmentionably worse. The hatefulness of this
practice is compounded by the fact that there is often no effective defense.
What can the most learned reader do when he is referred to a source
which he cannot reach without organizing a scholarly safari? There is
no clear and consistent way to distinguish useless references from useful
ones which, however remote their sources may be, are often invaluable
to specialists. But there are a few pernicious practices in citation which
are obviously illegitimate, and easily avoided. Does an author cite the
same sources over and over again with minor variations for the sake of
appearance? Does he use multiple citations, inextricably entangled in a
single note? Does he cite a primary source but quote a transcription from
a secondary one? Does he supply a partial citation, which is enough to
demonstrate his erudition but not enough to locate the source? These are
merely a few of many forms of scholarly malpractice in the construction
of references.

A third form of a pedantical ad verecundiam consists in the use of
quotations, which are often employed for forensic rather than empirical

12. Richard M. Zaner, "An Approach to Philosophical Anthropology," Philosophy and
Phenomenological Research 27 (1966): 55-68.

13. Kenneth R. Andrews, Drake's Voyages (New York, 1967), p. 67.

14. Marc Bloch, Strange Defeat (London, 1949), p. 27.

15. Richard Whately, Logic, p. 208.

FALLACIES OF SUBSTANTIVE DISTRACTION

287

purposes. In historical scholarship, this practice is widespread among
young scholars, who have a way of articulating a thesis in series of quota-
tions from older scholars and original sources — quotations which are
strung together like beads on a necklace, with a few connections of their
own invention. Their own best statements are sometimes buried in the
notes, where nobody can find them. As historians gain maturity, they
tend to become more assertive in their own right. But the habit is not
easily broken. 16

A variant upon this pedantic impropriety is the tendency of some
historians to cast their own doubtful interpretations in the form of attribu-
tion. Manning Clark, in his Short History of Australia ( New York, 1 963 ) ,
argues that "the life of the lower classes was all riot, revelry and drunken-
ness." 17 This dubious factual statement is accompanied by a more dubious
causal explanation, which is in turn supported by a most dubious method
of attribution. "Some observers attributed such behaviour to the dis-
crepancy between the sexes, especially amongst the convicts," Clark
wrote. 18 This, I think, is an argument ad verecundiam. It is certainly true
that there was a skewed sex ratio in Australia during the period under
discussion. But a causal connection cannot be sustained by attribution to
"some observers." Such interpretations are not properly theirs, but the
historian's and he must make an empirical case for them.

A fourth form of argument ad verecundiam consists in a thesis
which is sustained by the length of its exposition. Richard Whately com-
plained two centuries ago that "a very long discussion is one of the most
effective veils of fallacy. ... A fallacy which when stated barely . . . would
not deceive a child, may deceive half the world if diluted in a quarto
volume." 19 Many readers of Arnold Toynbee's A Study of History may
have been persuaded principally by the monumental proportions of the
work. Toynbee surely did not design this effect, but the results are the
same as if he had deliberately done so. The sheer pedantic bulk of his
many volumes is equally unnecessary to the articulation of his thesis, and
unpleasant in its irrational effect upon an awed and ignorant public. At
one point, Toynbee mentions "the Tarsian Jewish apostle of Christianity in
partibus inftdelium." He means Paul. In this "ornate alias," as it has been
called, nine pedantic words do the work of one. D. C. Somervell managed

16. For an example, see Donald Southgate, "The Most English Minister . . ." The
Policies and Politics of Palmerston (New York, I960), pp. 390, 392, 394, xxi, xxii, xxiii,
xxviii, 6, 8, 9, 10, 12, 13, passim; and see also a review of E. P. Thompson, The Making
of the English Working Class in the Economic History Review, 2d ser. 18, (1965):
632-43, for a discussion of the use and abuse of quotations by Thompson.

17. P. 73.

18. P. 73.

19. Whately, Logic, p. 151.

288

ARGUMENT

to shrink Toynbee's twelve volumes to two, largely by striking out the
adjectives. Overwriting of this sort is objectionable not merely for stylistic
reasons but also because of its rhetorical effect upon uncritical readers.

Isaac Disraeli, in his collection of literary curiosities, describes an
eccentric gentleman who invited a number of authors to dinner and seated
them in the order of the thickness of their publications. Precisely the
same bias operates when thick, square, quarto books attract a quantum
of enthusiasm out of all proportion to everything but their size. An ex-
ample is a fourteen-volume study of The British Empire Before the
American Revolution, by Lawrence Henry Gipson. The author is treated
with extreme deference by his colleagues in colonial history, partly be-
cause he has written an extraordinarily long book and partly because he
has lived an extraordinarily long time. 20 A more critical approach would
surely be more appropriate and also more complimentary to Gipson,
who is a good scholar and a tough old man. There are many method-
ological virtues in his work, and much that is useful as well. But there
are also serious deficiencies which ought to be sorted out by critical re-
viewers. Instead, his well-meaning colleagues tend to avoid any criticism
at all and yield to an unintended double ad verecundiam in their deference
to the length of the work itself and the length of the author's life. If con-
siderations of "taste" are to be consulted, then surely there is nothing as
tasteless as unqualified encomia to old men and big books.

A fifth form of a pedantical ad verecundiam is an appeal to detail
or specificity. The British philosopher, A. C. Bradley, has commented at
some length upon the fact that "we generally, it is true, take forcible de-
tail and strong particularity as a sign of fact." 21 Historians, especially, by
the nature of their discipline, are prone to this prejudice. There can, of
course, be no empirical proof without particularity, but there is often
particularity without proof. The reader might examine Macaulay's memo-
rable interpretations of Jeffreys, and Marlborough, and Penn, all of
whom he deeply disliked. His descriptions gain their argumentative power
and literary effect from a richness of controlled detail. But recent students
of these three men, as well as of other personages who became the
antagonists in Macaulay's work, have demonstrated that much of the
detail is false, misleading, or merely irrelevant to the point at issue. Par-
rington was another, of many great historians, who had the same extra-
ordinary gift, and who often misused it.

Sixth, there are arguments ad verecundiam in the form of an appeal
to the authority of mathematics. There is an anecdote, undoubtedly

20. A specimen of this devotional literature is Richard B. Morris, "The Spacious Empire
of Lawrence Henry Gipson," William and Mary Quarterly 24 (1967) : 169-89.

21. A. C. Bradley, The Principles of Logic, 2d rev. ed., 2 vols. (London, 1922), 1:75.

FALLACIES OF SUBSTANTIVE DISTRACTION 289

apocryphal, which may nevertheless serve as a hypothetical illustration.
It has often been said that Diderot once was asked in the court of Cather-
ine II of Russia to debate the existence of God with a mathematician
named Euler. The latter allegedly began by saying:

_a^JT_ = x
n

To this, he added, "Done Dieu existe. Repondez!" According to the
story, Diderot was so unnerved that he blushed and stammered and fled
for France.

It is unlikely that this tale is true. Diderot was not merely a great
controversialist but a good mathematician too. He would not have been
flustered by so trivial a forensic tactic. But true or false, this anecdote
is accurately descriptive of a common attitude among mathematical
illiterates. These unfortunate people are not merely untaught, but un-
teachable — until the fear of numbers is somehow gotten out of them.

The converse of this fallacy consists in the categorical rejection
of mathematical symbols. G. B. Shaw appears to have entertained a
great contempt for numbers. He wrote,

I somehow distrust mathematical symbols. I remember at school a plausible
boy who used to prove to me by algebra that one equals two. He always
began by saying, "Let x equal a." I saw no great harm in admitting that;
and the proof followed with rigorous exactness. The effect was not to
make me proceed habitually on the assumption that one equals two, but to
impress upon me that there was a screw loose somewhere in the algebraic
art, and a chance for me to set it right some day when I had time to look into
the subject. And I feel bound to make the perhaps puerile confession that
when I read Jevons's Theory of Political Economy, I no sooner glanced at
the words "let x signify the quantity of commodity,' than I thought of the
plausible boy, and prepared myself for a theory of value based on algebraic
proof that two and two make five. 22

There is something of this attitude in the response of historians
to the use of mathematics in their discipline by econometricians and
demographers. They might consider the reply to Shaw which was made
by a friend of Jevons', Philip H. Wicksteed, who wrote,

Mr. Shaw's youthful experience about x and a are so highly instructive that
I cannot refrain from dwelling upon them for a moment. His friend induced
him to "let x = a," and Mr. Shaw — not expecting that x would take any
mean advantage of the permission — granted the request. But he did not
understand that in letting x — a he was also letting xt — a = 0, and the
proof (of the proposition, 2=1) that "followed with rigorous exactness,"

22. Philip H. Wicksteed, The Common Sense of Political Economy, rev. and enl. ed.,
2 vols. (London, 1948), 2:726.

290

ARGUMENT

assumed that x — a did not equal zero. Mr. Shaw arrived at the sapient con-
clusion that there was "a screw loose somewhere" not in his own reasoning
powers, but — "in the algebraic art"; and thenceforth renounced mathe-
matical reasoning in favour of the literary method which allows a clever man
to follow equally fallacious arguments to equally absurd conclusions without
seeing that they are absurd. This is the exact difference between the mathe-
matical and literary treatment of the pure theory of political economy. 23

Finally, there is a common form of the fallacy of argument ad
verecundiam, which consists in the authority of the printed page over
ignorant minds. In most of us, there is a little of the legendary Caspar
Milquetoast, who believed anything he saw in writing — the blurb on
a book jacket, ballyhoo on a patent-medicine bottle, signboards, and
skywriting. Every teacher who has ever used a textbook has surely met
this form of error in some of his students.

There is also a counter fallacy which, according to Marc Bloch,
was popular among IhePoilus of the 1914-1918 war.

The prevailing opinion in the trenches [he wrote] was that anything might
be true, except what was printed. . . . The role of propaganda and censorship
was considerable, but in a way exactly the reverse of what the creators of
these institutions expected of them. . . . The men put no faith in newspapers,
and scarcely more in letters, for these, besides arriving irregularly, were
thought to be heavily censored. From this there arose a prodigious renewal
of oral tradition, the ancient mother of myths and legends. Wiping out by-
gone centuries by a daring stroke, beyond the wildest dream of the boldest
experimenters, governments reduced the front-line soldier to the means of
information and the mental state of olden-times, before journals, before news
sheets, before books. 14

There is perhaps a similar tendency today, among a certain
portion of the younger generation, to disbelieve anything which they
find in print, particularly on the subject of drugs, education, and Viet-
nam, unless it is the purple prose which oozes from their own hectograph
machines. But these same students are apt to be extraordinarily credulous
of any fact communicated to them by electronic means. This form of
error is not the antithesis of the Caspar Milquetoast syndrome, but its
complement. Both habits of thought are equally inimical to rational
intelligence and disciplined creativity.

£»> The fallacy of argument ad hominem occurs in many dif-
ferent forms, all of which serve to shift attention from the argument
to the arguer. Among its more common varieties are, first, the abusive

23. Ibid., 2:733.

24. Marc Bloch, The Historian's Craft (New York, 1953), pp. 107-8.

FALLACIES OF SUBSTANTIVE DISTRACTION 291

ad hominem, which directly denounces an opponent. The classic
example, perhaps apocryphal, is a note passed from one desperate
lawyer to another: "No case; abuse plaintiff's attorney."

Abraham Lincoln did this once, with the adversarial roles reversed.
Unhappily, he got away with it. Lincoln's law partner, William H. Hern-
don, remembered:

In a case where Judge [Stephen T.] Logan — always earnest and grave — op-
posed him, Lincoln created no little merriment by his reference to Logan's
style of dress. He carried the surprise in store for the latter, till he reached
his turn before the jury. Addressing them, he said: "Gentlemen, you must be
careful and not permit yourselves to be overcome by the eloquence of counsel
for the defence. Judge Logan, I know, is an effective lawyer. I have met him
too often to doubt that; but shrewd and careful though he be, still he is
sometimes wrong. Since this trial has begun I have discovered that, with all
his caution and fastidiousness, he hasn't knowledge enough to put his shirt
on right." Logan turned red as crimson, but sure enough, Lincoln was cor-
rect, for the former had donned a new shirt, and by mistake had drawn it
over his head with the pleated bosom behind. The general laugh which fol-
lowed destroyed the effect of Logan's eloquence over the jury — the very
point at which Lincoln aimed. 29

A second variety of argument ad hominem is circumstantial. It
consists in a suggestion that an opponent's argument is merely a reflection
of his interest. Adlai Stevenson attacked the arguments of the Repub-
lican party in 1952 with the assertion that it was "out of patience, out of
sorts, and out of office." 2 *

Third, there are associative ad hominem arguments, which attempt
to undercut an opponent by reference to the company he keeps. There
is an example, in doggerel, from English politics:

If the Devil has a son,
It is surely Palmerston.

A pictorial specimen from American politics was the attempt by Sen-
ator Joseph McCarthy and his staff in 1950 to end the political career
of a consistent and courageous critic — Senator Millard Tydings of
Maryland — by distributing a doctored photograph showing Tydings
tete-a-tete with a prominent American Communist, Earl Browder. The
attempt was successful. 27

Fourth, there are arguments ad hominem in the form of tu quoque
("you too"), in which it is suggested that an opponent has sometimes
held the view which he now opposes, or that he has adopted the practice

25. Paul M. Angle, ed., Herndoris Ufe of Lincoln (New York, 1965), p. 291.

26. Adlai Stevenson, Major Campaign Speeches (New York, 1953), p. 4.

27. Richard Rovere, Senator Joe McCarthy (New York, 1959), pp. 160-61.

292

ARGUMENT

which he new condemns, or that his argument applies to himself as well
as to his opponent. A graffito, scrawled upon the wall of a New York
subway station, read:

God is Dead: Nietzsche
To this, some subterranean pilgrim made the reply:
Nietzsche is Dead: God

One of the great ad hominists of all time was Dr. Johnson. Oliver
Goldsmith was driven to complain that "there is no arguing with John-
son: for if his pistol misses fire, he knocks you down with the butt end
of it." 28 Johnson employed this brutish tactic not merely in the form
of argument ad hominem, but ad feminam as well. It is reported that
"Once, upon hearing a lady from the provinces complain of how
she disliked London because there her fingernails were always dirty,
he remarked, 'Perhaps, Madam, you scratch yourself.' " 29 In a more
famous encounter, he was conversing with Mrs. Catherine Macaulay,
a female historian of radical Whiggish principles, and an enthusiast
for the American Revolution — which is to say that by her very existence
she infuriated her bearish antagonist in seven different ways. When Mrs.
Macaulay argued boldly for the equality of mankind — in the presence of
her servants — Dr. Johnson replied by asking her to allow a footman to
sit down beside her. 30

Once Dr. Johnson got as good as he gave, from another female
of his acquaintance. In conversation with Miss Seward, a lady poet
from his native Lichfield, he asserted, "I am willing to love all mankind,
except an American." Then, according to Boswell,

his inflammable corruption bursting into horrid fire, he "breathed out
threatenings and slaughter;" calling them, "Rascals — Robbers — Pirates;" and
exclaiming, he'd "bum and destroy them." Miss Seward, looking to him with
mild but steady astonishment, said, "Sir, this is an instance that we are al-
ways most violent against those whom we have injured." — He was irritated
still more by this delicate and keen reproach; and roared out another tre-
mendous volley which one might fancy could be heard across the Atlantick.
During this tempest I sat in great uneasiness, lamenting his heat of temper;
till, by degrees, I diverted his attention to other topicks 31

Arguments ad hominem sometimes run in reverse, in which case
they are, I suppose, arguments de homine. The French historian Al-
phonse Aulard made a career of attacking Hippolyte Taine. Aulard

28. lames Boswell, The Life of Dr. Johnson, p. 361, Oct. 26, 1769.

29. D. Judson Milburn, The Age of Wit (New York, 1966), p. 24.

30. Boswell, The Life of Dr. Johnson, p. 660, 1776.

31. Ibid., p. 806, April 15, 1778.

FALLACIES OF SUBSTANTIVE DISTRACTION 293

even went so far as to assert that "a candidate for a history degree at
the Sorbonne would not make a good impression if he quoted Taine
as an authority." 32

There are many arguments ad hominem in historiographical ex-
changes. The nature of this fallacy is, I hope, sufficiently clear not to
require additional examples. If anybody is interested, he might consult
the "Communications" section of the American Historical Review, in
which ad hominem arguments are exchanged as regularly as a ball in
a tennis match. But an ad hominem debate is unlike tennis in one
respect — it is a match which everybody loses: players, referees, spec-
tators, and all. 33

The fallacy of argument ad crumenam makes money into
a measure of truth and right. A vulgar example is the proverbial retort
which finds a regrettable popularity in this republic — "If y'r so smart,
why ain't ya rich?" I had not expected to find an instance of this error
in the work of academic historians, whose economic status is a sufficient
discouragement. Indeed, among threadbare academic gentlemen, the
counterfallacy of an argument to poverty is perhaps more popular,
except when it is used by college presidents, as it actually was by
Harvard's Charles W. Eliot, as an excuse to keep down faculty salaries.

But an unexpected example of argument ad crumenam turned up
in the work of President Eliot's kinsman — Samuel Eliot Morison. In
a plea for the excellence of old-fashioned narrative writing, Morison
argued that narratives were better than monographs, partly because
they were worth more money to their authors.

The tremendous plowing up of the past by well trained scholars is all to the
good, so far as it goes [he wrote]. Scholars know more about America's past
than ever; they are opening new furrows and finding new artifacts, from
aboriginal arrowheads to early 20th century corset stays. But they are heap-
ing up the pay dirt for others. Journalists, novelists and free-lance writers
are the ones that extract the gold, and they deserve every ounce they get
because they are the ones who know how to write histories that people care
to read. What I want to see is a few more Ph.D.'s in history winning book-
of-the-month adoptions and reaping the harvest of dividends. They can do
it, too, if they will only use the same industry at presenting history as they
do in compiling it. 34

32. Pieter Geyl, Encounters in History (New York, 1961), p. 132. Aulard's attack ap-
peared at greatest length in Taine: Historien de la Revolution Francaise (Paris, 1907).

33. See, e.g., The American Historical Review 73 (1968): 996, 1710, passim.

34. Samuel Eliot Morison, "History as a Literary Art," Old South Leaflets, series 2,
number 1 (n.d.): 5.

294

ARGUMENT

The reader will note that an argument ad crumenam is grafted on to
an argument ad populum. Morison has other arguments for history as
literary art, most of which are equally mistaken. He meant well, and
the stylistic deficiencies of much professional prose are a common
scandal. But this argument for the merits of literary history might be
translated precisely into the vulgar terms of our first example: "If
y'r so smart, why ain't ya rich?"

$»» The fallacy of argument ad baculum (literally, to a big stick)
is an appeal to physical force, actual or implied, in order to sustain an
argument — a crude tactic not commonly employed in American aca-
demic disputes today, unless students are the disputants. Tempers have
flared to a fever heat in many a fierce interpretative exchange. But
never, to my knowledge, have historians of slavery attempted to settle
the profitability question at twelve paces; nor have embattled Turn-
erians and anti-Turnerians ever conducted a gouging contest on a
prairie campus. No critic of the New Political History has been defenes-
trated from Pittsburgh's tower of learning; nor has a graduate student
ever actually been pressed to death beneath a folio set of the Monumenta
for standing mute on a doctoral examination.

But it was not always so. The most powerful piece of historical
writing in Western culture ends, characteristically, with an omnibus
threat of bodily injury and spiritual harm to all would-be revisionists.
The reader of the Book of Revelation, 22:18-19, is solemnly warned:
"I testify unto every man that heareth the words of the prophecy of
this book, If any man shall add unto these things, God shall add unto
him the plagues that are written in this book: And if any man shall
take away from the words of the book of this prophecy, God shall take
away his part out of the book of life, and out of the holy city, and from
the things which are written in this book."

Today a soft sell is thought to be more effective. But there have
been many disagreeable instances of an argument ad baculum in the
great disturbances which have disrupted the peace of the world in the
past few generations. In Japan, during the reign of the present emperor,
a scholar who doubted the accuracy of the standard genealogical history
of the imperial family was threatened with assassination. In Soviet
Russia, and other Communist nations, more than a few chapters of
historiographical revisionism have ended in the rattle of machine-gun
fire. German historiography has perhaps suffered more severely in this
respect than that of any other nation. Sixty years ago, its historians set
standards for the world. Thirty years ago, its great centers of learning

FALLACIES OF SUBSTANTIVE DISTRACTION 295

were purged of every vestige of open anti-Fascism. Today in historical
scholarship, Germany is a stagnant backwater. This great reversal can-
not be attributed solely to the nihilistic violence of Nazism; other agents
have contributed to the effect. Nevertheless, the decline of the German
universities since 1932 testifies to the dangerous power of an argument
ad baculum, not merely over the specific victim, but over everybody in
the vicinity.

These examples are not likely to be very disturbing to liberal
Americans — Fascist beasts and fanatic Bolsheviks are supposed to mis-
behave. But there have also been many regrettable examples in modern
America. James W. Silver, a historian in Mississippi, had the temerity,
a few years ago, to speak openly and honestly of the culture in which
he lived. He was threatened forthwith with bodily violence by the people
who set the tone of public discussion in that state. There are such prim-
itives living in the nooks and crannies of every civilized society. But
society in Mississippi is so organized as to magnify their power and
their opportunities to act. More recently, in many universities through-
out the western world, radical students have disrupted classes, damaged
buildings, destroyed books, and injured people, and committed many
outrages when instructors made statements with which they disagreed.
In Massachusetts recently, students answered an unpopular speaker by
thrusting a banana in his mouth. In California, other people have been
killed in these violent disturbances. Three years ago, when I began work
on this book, I had planned to omit argument ad baculum, on the ground
that it occurred too infrequently in American universities to warrant
mention. How wrong I was! Today, it constitutes a more immediate
threat to rational processes than any other error which I have included.

What is a rational man to do, in the face of an appeal ad baculum?
Knock-down arguments, alas, must be overcome not with a syllogism
but a stick. Liberty and order are the prerequisites of reason. Let us
hope that the senseless and suicidal misbehavior of a small number of
immature students, and of a still smaller number of irresponsible
adults who have urged them on, will awaken the great moderate mass
of free men in this republic to protect and defend our rare and happy
heritage of freedom and stability. Let us have the courage, patience, and
wisdom to enforce restraint (without repression) upon our erring chil-
dren. Then, only then, can the dialogue of reason continue.

There is, by the way, a counterfallacy which consists in an appeal
not to a big stick but to a little one. Americans who pride themselves
upon a proverbial sympathy for the underdog are perhaps particularly
vulnerable. There is surely no a priori reason for taking up the cause of
the underdog, who may be a dirtier dog than the top dog. Critics of

296

ARGUMENT

the war in Vietnam have raised the cry of "Bully" against American
intervention. But right does not vary with might — even inversely.

The fallacy of argument ad temperantiam is an appeal to
moderation, on the apparent assumption that truth, in Burke's phrase,
is always a "sort of middle." In academic scholarship it commonly occurs
in two forms. The first is stylistic; the second, substantive.

The prevalance of a prejudice toward stylistic moderation in aca-
demic historiography explains the monumental dullness of its mono-
graphic literature. The problem is not that historians do not know how
to write, but rather that they are actively discouraged from indulging
in any stylistic practice which might be interpreted as literary intemper-
ance. Historians write dull books not because they are dull fellows, but
because they have formed the stupid habit of confusing dullness with
detachment. A clear example of an excess of stylistic moderation is the
good, gray, Cambridge History of British Foreign Policy, or the Cam-
bridge Modern History. An editor of these excruciating works, the Eng-
lish diplomatic historian A. W. Ward, is alleged to have revised one
chapter on the ground that "It's a bit lively." 35

An exceedingly lively American scholar, Charles A. Beard, may
have operated upon the same assumption in his most important work,
An Economic Interpretation of the Constitution. Several scholars have
pondered an odd fact about that work. Eric Goldman observes,

For a man like Charles Beard, who could write with either a shillelagh or a
stiletto, the Economic Interpretation was a triumph in dullness. Not a single
departure from catalogue organization, not a single bright sentence, en-
livened the book, which carefully described itself as an "arid survey." Max
Lerner, who believes that the Economic Interpretation influenced his own
thinking enormously, has described the volume's strategy of flatness. "Beard
must have had a premonition of the desperate resistance he would run into,"
Lerner commented. ". . . It is almost as if the author had set out with a
deliberate severity to strip the book of every adornment, on the theory that a
plain woman would be less suspected of being a wanton than an attractive
one." 38

If liveliness is thought to be unbecoming in a monograph, enthu-
siasm appears to be unforgivable. One of Beard's most outspoken critics,
Robert Brown, failed to bend before his colleagues' bias toward moder-
ation and published several volumes in a spirit of furious revisionism,
which are sometimes condemned not for their substance but for their

35. S. C. Roberts, Adventures with Authors (Cambridge, 1966), pp. 112-13, quoted in
G. R. Elton, The Practice of History (New York, 1967), p. 108.

36. Eric Goldman, Rendezvous with Destiny (New York, 1952), p. 153.

FALLACIES OF SUBSTANTIVE DISTRACTION ?97

style. A conservative scholar has written, "What mars Professor Brown's
important contributions to our understanding of the late colonial period
is a certain belligerence toward fellow workers in this field." 37 I hold no
brief for Brown, who does indeed make himself more than a little
disagreeable by a strident, self-righteous style. But that aspect of his
work has nothing whatever to do with its interpretative merit.

A second form of the fallacy of argument ad temperantiam con-
sists in substantive moderation. N'ayez pas de zele," said Talleyrand.
Many historians have made this mot into a motto. Ranke was one of
them. Beard sneered that "he could write of popes in a manner pleasing
to both Catholics and Protestants of the upper classes." That fact was
owing not merely to an elitist common denominator but to an excess
of substantive moderation. Ranke, I think, sometimes imagined that
he had unlocked the inner secret of objectivity, when he had merely
found a middling subjectivism.

Another example was Macaulay, who condemned historians who
believed that to be impartial was "to intersperse in due proportion
epithets of praise or abhorrence." Macaulay himself operated on a
different assumption. To be impartial was, for him, to intersperse, in
due proportion, epithets of abhorrence for all extremes. A recent work
has analyzed the importance of "middlingness" in his thought. 38

The fallacy of argument ad antiquitam is an illegitimate
appeal to ages past in order to justify acts present or future. Jeremy
Bentham called it the "Chinese argument" in his Handbook of Political
Fallacies, which is both inaccurate and unfair in its implication that
this fallacy is a logical disorder to which orientals are especially suscep-
tible. Bentham's own hilarious examples suggest that an argument ad
antiquitam has perhaps been brought to its highest level of refinement by
lily-white Anglo-Saxon gentlemen on both sides of the Atlantic Ocean.
Constitutional law in both England and America might be conceived as
one prolonged and preposterous argument to antiquity by inscrutable
occidentals in flowing judicial robes. "Men love old truths," said Billy
Herndon, as he traveled the circuit in frontier Illinois, and they love
old errors, too. There is scarcely a corner of the world in which men
do not, in some degree, bow down before absurdities inherited from
their ancestors.

37. Lawrence Henry Gipson, The British Empire Before the American Revolution, vol.
13 (New York, 1967), p. 453.

38. Vincent E. Staizinger, Middlingness: Juste Milieu Political Theory in France and
England, 1815-1848 (Charlottesville, Va., 1965).

298

ARGUMENT

An argument to age is not merely irrelevant to many of the
issues which it has been used to defend. It is also, in its usual form,
inconsistent with itself. Let us assume for the moment that age is wiser
than youth, presumably because of its experience. If our ancestors were
alive today, they would be venerable indeed, and their opinions should
carry considerable weight. But they are not alive, an obvious fact which
is often forgotten by ancestor-worshiping fanatics. Now let us add a
second premise which is also necessary in an argument to age — namely,
that men can and do learn from the experience of others, and that there
is a kind of collective experience which customarily attaches to all man-
kind, and which is commonly called its history.

Now, when our ancestors lived, they were as young as we are
(younger on the average), and the world was younger still. The collec-
tive experience of mankind was less extended in their time than ours.
Their opinions, therefore, partook more nearly of the youth of man-
kind than ours do. As the date of their opinions is distant from the
present, so in the same proportion are their opinions less mature than
our own. An eighteenth-century idea, in this sense, is not two centuries
older than a twentieth-century idea, but two centuries younger. If the
premises of an argument to age are granted, then the distance of an
ancestor's opinions from the present should be in the same degree a
presumption against their validity rather than for it. 39

It is possible, of course, to make a different kind of argument to
ancestors, which is logically more tenable, if empirically more prepos-
terous. Such an argument might begin with Henry Adams's assumption
that there is an equivalent to the second law of thermodynamics in his-
tory, so that the energy and wisdom of all mankind are always running
downhill. But this is an argument to youth, and not to age.

However absurd these thoughts may be, many intelligent historians
have followed them. In another work, I have argued at some length that
Henry Cabot Lodge, a Harvard Ph.D. in history and a competent
historian in his own way, did this, with profoundly deleterious results
for his own career. He attempted to apply the precepts of his great-
grandfather in a political world where practices were very different. 40

In our own time, it is a rare political proposal which is not, in
some fashion, legitimized by an out-of-context quotation from the
Founding Fathers. Consider, for instance, the rhetorical raids which
have been made upon the writings of Thomas Jefferson. His works

39. William P. Montague, The Ways of Knowing, or the Methods of Philosophy (New
York, 1958), p. 44.

40. David Hackett Fischer, "Founding Fathers and Great Grandsons" (in Life and
Letters of George Cabot, by Henry Cabot Lodge forthcoming).

FALLACIES OF SUBSTANTIVE DISTRACTION 299

have been ransacked by Democrats and Republicans, liberals and con-
servatives, radicals and reactionaries, New Englanders and Southerners,
to sustain elitism and equality, capitalism and socialism, states' rights
and interventionism, isolationism and internationalism, rationalism and
romanticism, atheism and Christianity, agrarianism and urban develop-
ment. He has been quoted at length by Earl Browder in defense of
Communism, and by Ezra Pound in the cause of Fascism; by Sukarno
in the interest of "guided democracy," and even by Ho Chi Minh in the
name of Vietnamese nationalism. All of this, of course, has often been
abetted by historians, who have too often allowed themselves to be
seduced into an essay called "Thomas Jefferson survives," or some
such thing. Among the perpetrators of these wretched little essays are
some of the most able historians in the Republic — Ralph Henry Gabriel,
Henry Steele Commager, Julian Boyd, Dumas Malone, Charles A.
Beard, C. M. Wiltse, A. M. Simons, and many others. 41

£*> The counterjallacy of argument ad novitam is an appeal to
modernity, or novelty, or youth. Its popularity among historians is per-
haps partially explained by Marc Bloch's anecdote about another
great scholar.

I had gone with Henri Pirenne to Stockholm. We had scarcely arrived, when
he said to me: "What shall we go to see first? It seems that there is a new
city hall here. Let's start there." Then, as if to ward off my surprise, he
added: "If I were an antiquarian, I would have eyes only for the old stuff,
but I am an historian. Therefore I love life." 42

Whatever the explanation may be, arguments ad novitam appear in
several forms in historical writing. One of them is the unfortunate habit,
which historians have formed and which the public has encouraged, of
assuming that special significance attaches to a subject if it can be
proved to be chronologically first in some respect. David Marshall
Lang, author of a fine biography of Alexander Radischev, argues in his
title for the significance of his protagonist by calling him The First Rus-
sian Radical (London, 1959). But if "radical" means "a person who
favors rapid and sweeping changes in laws and methods of government,"
as Webster defines it, then there were surely Russian radicals before
Radischev, and others after him. It could be argued that each of them
was "first" in some respects but not in others. However that may be,
Radischev retains a historical significance for the kind of radical that he

41. See, generally, Merrill Peterson, The Jeffersonian Image in the American Mind (New
York, 1960).

42. Bloch, The Historian's Craft, p. 43.

300

ARGUMENT

was and for the radical things that he did. The question of firstness is a
minor matter, of much less importance.

Another common historiographical form of an ad novitam is the
excessive weight which often attaches to the most recent interpretation
of any given subject. In one sense, this is the very opposite of firstness,
for the blue ribbon is awarded to the book in last place. But, looking
backward, it is the first book — the newest and most novel. "The judg-
ment that earlier accounts are untrue, or at best inaccurate, is common
to historians in all periods," Frederick Teggart complained in 1925.
"The remarks of Thucydides on Herodotus, of Polybius on Timaeus, of
Lucian on Ctesias, are typical of ancient historiography. In modern
times, the same attitude has been maintained. Macaulay and Froude
are the butt of every novice. Round calls Freeman "a superseded fossil." 43

It is true that the best studies of certain historical problems are
often the most modern studies. But there is no necessary causal or logical
connection between modernity and excellence. There are many cases in
point, within my own field. The best survey of the political history of
the American Republic in the 1 790s is still Richard Hildreth's account,
published more than a century ago. The best general history of the Jef-
fersonian era is still Henry Adams's, despite many flaws. The best single
book on the so-called Age of Jackson is still Tocqueville's, by a long
shot. There are many other examples. The best history of medieval
Florence is still the great work of Davidsohn; the best history of the
Spanish Inquisition, in my opinion, is still Llorente's; the best history of
the origins of the Seven Years' War is still Ranke's; the best military
history of the Crimean War is surely Kinglake's. A competent scholar
has written that there are "aspects of the history of pre-Columbian Amer-
ica in which the last, best work is still the first serious work — Alexander
von Humboldt's extraordinary Vues de Cordilleres et Monuments des
Peuples Indigenes de I'Amerique (1814)." 44

There is also a good deal of the fallacy of argument ad novitam in
the absurd child worship of the late 1960s — in its juvenile fashions and
fads, and in the infantile conceit that nobody over thirty can ever be
trusted. Surely an inverse age requirement for integrity is as absurd as
Noah Webster's idea that franchise should be restricted to people over
forty-five.

$»» The fallacy of argument ad consequentiam is an attempt to
prove or disprove a reasoned argument by reference to the consequences

43. Frederick J. Teggart, Theory and Processes of History (Berkeley, 1962), p. 13.

44. Leo Deuel, Conquistadors Without Swords: Archaeologists in the Americas (New
York, 1967).

FALLACIES OF SUBSTANTIVE DISTRACTION 301

which flow from its acceptance or rejection. One example is the verdict
which an English historian, G. R. Elton, passed upon Erik Erikson's
psychoanalytic interpretation of Martin Luther. "I cannot feel that the
much-praised Freudian effort of E. H. Erikson, Young Man Luther: A
Study in Psychoanalysis in History (London, 1959), contributes any-
thing of value to an understanding of Luther or his age," Elton wrote,
"In so far as it has been responsible for John Osborne's play, it may
even need condemnation."* 5

Elton is, of course, entitled to his opinion, which must be regarded
with respect, particularly when it refers to European history in the
sixteenth century. But his second sentence is surely illegitimate. Maybe
Mr. Erikson's essay was responsible, in some degree, for John Osborne's
play, in the sense that the playwright may have derived many of his
ideas and most of his facts from the psychiatrist. (I don't know that this
is the case.) And maybe Osborne's play itself calls for condemnation
as a historical work. Its interpretation of Luther is at least as strained as
it alleges the Lutheran bowels to have been. But one cannot reason from
these premises to the conclusion that Erikson's monograph should be
condemned for the deficiencies of an unfortunate melodrama which may
have been based upon it.

Other historiographical examples of an argument to consequences
appear in attempts by conservative or chauvinist historians to justify
the United States' war with Mexico. A nineteenth-century German
scholar, Hermann von Hoist, believed that Polk's policy was a Good
Thing, because it caused the civilization of California, which was a
Very Good Thing.

In the hands of Mexico [he wrote] California was not only as good as lost
to civilization, but it also lay exposed, a tempting prey, to all the naval and
colonial powers of the world. ... In whatever way the ethics of ordinary
life must judge such cases, history must try them in the light of their results,
and in so doing must allow a certain validity to the tabooed principle that
the end sanctifies the means. Its highest law is the general interest of civil-
ization, and in the efforts and struggle of nations for the preservation and
advancement of general civilization, force not only in the defensive form,
but also in the offensive, is a legitimate factor. 46

Von Hoist's interpretation is doubly objectionable, on ethical and
logical grounds. The end may or may not justify the means, but it can-
not sanctify them. Even assuming the validity of Von Hoist's absurd
premise that Mexico's loss was civilization's gain — an assumption which

45. Elton, The Practice of History, p. 25.

46. Hermann von Hoist, The Constitutional and Political History of the United States,
8 vols. (Chicago, 1881) 3:268-73.

302

ARGUMENT

is not likely to survive a cultural comparison of Los Angeles and
Mexico City — it does not follow, that a quality which attaches to an effect
is transferable to the cause, or that a line of reasoning is valid or invalid
because we happen to approve or disapprove of its practical conse-
quences.

A superpatriotic American historian has taken a similar position
on the same issue. Samuel Flagg Bemis commemorated the expansion-
ism of President Polk with a combination of an argument ad conse-
quentiam and an argument ad populum. To critics of the aggressive acts
which led to the Mexican cession, Bemis asked if they would wish to give
it back. "Notwithstanding all this," he wrote at the end of his account
of the aggression of the United States against its neighbor, "Notwith-
standing all this it would be well-nigh impossible today to find a citizen
of the United States who would desire to undo President Polk's diplo-
macy, President Polk's war, and the treaty of Guadeloupe Hidalgo."* 7

$»» The fallacy of argument ad nauseam might serve, tongue in
cheek, as a tag-name for a serious form of error, in which a thesis is
sustained by repetition rather than by reasoned proof. This strategy was
a favorite of Lewis Carroll's immortal Bellman, in The Hunting of the
Snark.

"Just the place for a Snark!" the Bellman cried

As he landed his crew with care;
Supporting each man on the top of the tide

By a finger entwined in his hair.

"Just the place for a Snark! I have said it twice:

That alone should encourage the crew.
Just the place for a Snark! I have said it thrice:

What I tell you three times is true." 48

The absurdity of this ludicrous device appears merely in its description.
But it should not be lightly dismissed. Its popularity among advertising
executives, public relations specialists, and professional propagandists is
sustained by the solid fact that it works. Recent research in psychology
has suggested something of the subliminal mechanisms by which repeti-
tion erodes critical resistance to the most absurd assertions.

Everybody, I suspect, has met this fallacy in various forms. An
example is provided by an American historian, Oscar Handlin. There is a

47. Samuel Flagg Bemis, A Diplomatic History of the United States, 4th ed. (New
York, 1955), p. 244.

48. The Complete Works of Lewis Carroll, p. 757.

FALLACIES OF SUBSTANTIVE DISTRACTION 303

salutary unwritten rule in the historian's code of professional ethics
that forbids him to review the same book more than once. But there is
no comparable taboo against his writing the same book as many times
as he likes. Handlin's bibliography is enormous. But his many titles are
mostly variations on a few themes.

The most prominent theme is the immigration of non-English
people to America, mostly since 1820, which Handlin has implicitly
elevated into the central "factor" in American history in his works. This
argument has not been received with enthusiasm by critics, one of
whom fairly complained that Handlin tended to ignore the contribution
of the Anglo-Saxon element to American culture. But Handlin does not
reply — he republishes. The result is a bibliography which reads a little
like the Bobbsey Twins series. The Uprooted is followed by Sons of the
Uprooted; Boston's Immigrants by Positive Contributions by Immi-
grants and Immigration as a Factor in American History. The same
arguments also appear in The Americans: A New History of the People
of the United States, which is about "the influence of migration upon the
people of the United States"; and in The American People in the Twen-
tieth Century, which might be called "Immigration as a Factor in
Twentieth-Century America," and Al Smith and His America, which is
about "Immigration as a Factor in Al Smith's America."

Similarly, Handlin has argued that the predicament of Negroes in
twentieth-century America is closely comparable to that of immigrant
groups. He has been sharply criticized for underestimating the duration
and intensity of race prejudice and the significance of race slavery in
America; and also for tending to mistake Negroes for Jews, and maybe
for German Jews — a good example of the culture-bound fallacy. Again,
Handlin responds by reiteration: Race and Nationality in American Life
(1957); The Newcomers: Negroes and Puerto Ricans in a Changing
Metropolis (1959); Fire Bell in the Night (1964); and many short
essays besides. The words of Tristram Shandy come to mind:

Tell me, ye learned, shall we for ever be adding so much to the bulk —
so little to the stock?

Shall we for ever make new books, as apothecaries make new mixtures,
by pouring only out of one vessel into another?

Are we for ever to be twisting, and untwisting the same rope? for ever
in the same track — for ever at the same pace?

Shall we be destined to the days of eternity, on holy-days as well as
working-days, to be shewing the relics of learning, as monks do the relics
of their saints — without working one — one single miracle with them? 49

49. Laurence Sterne, The Life & Opinions of Tristram Shandy, Gentleman, Modern
Library ed. (New York, 1950), pp. 355-56.

304 ARGUMENT

There are many similar distractive fallacies of argument: ad
misericordiam (an appeal to pity); ad odium (to hatred); ad super sti-
tionem (to credulity) ; ad modum (to gradualism, due measure, or pro-
portion) ; ad metum (to fear) ; ad superbiam (to snobbery, or pride) ; ad
invidiam (to envy), and others, without limit. Many of them are discussed
at length by Jeremy Bentham in his Handbook of Political Fallacies.

All of these forms of error are committed in historiographical as
well as forensic disputation. Consider, for instance, the historiography
of one great historical event — the French Revolution. A major compo-
nent of Michelet's interpretation is an argument ad misericordiam in the
name and interest of "the people," most of whom appear to have been
hungry widows and orphans.

Carlyle's The French Revolution is equally an argument ad odium.
His hatred is visited upon the alleged chaos of the Revolution itself, that
great madness, "World Bedlam," and also upon its greatest personality,
Robespierre, "seagreen Pontiff," "most consistent, incorruptible of thin,
acrid men." Even his virtues are converted into objects of loathing.

Foes and friends of the Revolution have both resorted to an argu-
ment ad super stitionem. In the vanguard of the former marched the pious
Abb6 Barruel, who thought that the Revolution was the dark spawn
of an unnatural union between the Incubus of Freemasonry and the
Bitch-Goddess Liberty. On the other side, the Christian socialist Buchez
interpreted the French Revolution as the institutionalization of Chris-
tianity. Some parts of his forty volumes read like an extended com-
mentary on the New Testament.

The biases which inform Tocqueville's Old Regime and the French
Revolution are a clear example of argument ad modum. Taine's history,
on the other hand, is an extended ad metum. That great historian — a
moderate without moderation, and a liberal without liberality — will long
be remembered for his hair-raising descriptions of insane rationalism
of the social contract, for his horror tales of the flashing bright blade of
the guillotine, and for his interpretation of the people as a great beast,
"I'animal primitif, le signe grimacant, sanguinaire, et lubrique." All of
these images are symbols not merely of hatred, but fear. Taine had
witnessed some similar sights in his own time — and he was afraid.

Arguments ad superbiam, in the form of class pride, fill the books
of conservative critics — Maistre and many since. As culture pride, the
same form of error appears in the works of Renan and Flaubert. As
race pride it is deeply rooted in Maurras, Gaxotte, and Bainville. On
the other side, arguments ad invidiam appear on occasion in a sequence
of socialist historians from Louis Blanc to Mathiez and Georges Lef ebvre.

FALLACIES OF SUBSTANTIVE DISTRACTION 305

By way of a conclusion, there is nothing to add on the dismal
subject of substantive distractions except a few tedious platitudes which
have undoubtedly suggested themselves already to every serious reader.
Instead, I shall end this chapter and the main body of the book by
cataloguing one final fallacy of distraction to which a fallacist himself
is especially susceptible. It is offered as a byword and a warning to
every reader who has responded in the wrong spirit to this work. The
fallacist's fallacy consists in any of the following false propositions:

1. An argument which is structurally fallacious in some respect
is therefore structurally false in all respects.

2. An argument which is structurally false in some respect, or
even in every respect, is therefore substantively false in its conclusion.

3. The appearance of a fallacy in an argument is an external sign
of its author's depravity.

4. Sound thinking is merely thinking which is not fallacious.

5. Fallacies exist independent of particular purposes and assump-
tions.

All of these propositions are profoundly wrong. Let us examine them
briefly. First, an argument which is fallacious in one of its parts is
not necessarily fallacious in others. All great historical and philosophical
arguments have probably been fallacious in some respect. But it is un-
likely that any extended argument has ever actually been fallacious in
all respects. Complex theses are great chains of reasoning. The fact that
one link in the chain is imperfect does not mean that other links are
necessarily faulty, too. If the argument is a single chain, and one link
fails, then the chain itself fails with it. But most historians' arguments are
not single chains. They are rather like a kind of chain mail which can
fail in some part and still retain its shape and function. If the chain mail
fails at a vital point, woe unto the man who is inside it. But not all
points are vital points.

Second, even if an argument is structurally fallacious in such a way
that it collapses as an argument, the conclusion may still be substan-
tively correct. Many false arguments have yielded true conclusions. An
eighteenth-century New England physician studied the origin of the
American Indians and arrived at the accurate conclusion that they were
of Mongol extraction. His reasoning was roughly as follows: Noah had
three sons: Japhet, Shem, and Ham. Everywhere, the children of Ham
must serve the children of Japhet. The Indians have no Negro slaves,
and therefore they must be the children of Shem. Thus, they are of
Mongol extraction. 50 There is nothing good to be said for this argument

50. Sir Lewis Namier, in Personalities and Powers (London, 1955), "Human Nature in
Politics," p. 6.

306 ARGUMENT

from either a logical or an empirical point of view. But its conclusion
is substantively valid, even though every one of its premises is false,
and the method of inference as well.

Third, a fallacy is not an emblem of depravity, or evidence of
deliberate deceit. We have already noticed this error in Jeremy Bentham's
Book of Fallacies. It is a common mistake. One of my colleagues ex-
amined an example of a fallacy which appears in an earlier chapter of
this book — a fallacy from the work of X, a distinguished American
historian. The colleague commented, "What you really mean to say,
is that A" is a Bad historian!" And his expression implied, "A Bad
Man, to boot." I meant no such thing. All examples of fallacies in this
book are drawn from the work of competent historians. Some are from
the work of great historians. None, to my knowledge, were deliberately
concocted to deceive a reader.

Fourth, the incidence of empirical and logical error in a work is
not an inverse measure of its excellence. There are no rules of reason
for distinguishing a shallow argument from a profound truth. There are
no axioms of logic which can measure the gift of creativity or wisdom.

Finally, forms of illogic are always relative to specific logical
assumptions and objects. They do not possess an independent existence.
Aristotle's "fallacy of accident" is dependent upon the assumption that
there are distinct differences between "essential" and "accidental" qual-
ities. But if one disbelieves in the existence of essences, then the fallacy
of accident has no meaning. Some of the fallacies in this book may be
as empty to some readers as the fallacy of accident is to me.

The thesis of this book is something distinct from all these errors.
It is a simple proposition that historians, and all men who seek to
think historically, tend to make certain assumptions in their work, and
that these assumptions have logical consequences which must be re-
spected. Every historical work has a logical dimension. Of course, it has
many other dimensions, too. If there is a logic of historical thinking,
there is equally a grammar and a rhetoric of historical expression. And
if there is a historiographical trivium, there is much more besides which
belongs to the realm of psychology, epistemology, ethics, and meta-
physics. But of all these many components of historical thinking, none is
more susceptible to control and refinement than the logical component.
Logic is not everything. But it is something — something which can be
taught, something which can be learned, something which can help us
in some degree to think more sensibly about the dangerous world in
which we live.

CONCLUSION

History is not only a particular branch of knowledge, but a
particular mode and method of knowledge in other branches.
— Lord Acton

Any serious attempt to answer the question "What is good history?"
leads quickly to another — namely, "What is it good for?" To raise this
problem in the presence of a working historian is to risk a violent
reaction. For it requires him to justify his own existence, which is
particularly difficult for a historian to do — not because his existence is
particularly unjustifiable, but because a historian is not trained to
justify existences. Indeed, he is trained not to justify tliem. It is usually
enough for him that he exists, and history, too. He is apt to be impatient
with people who doggedly insist upon confronting the question.

Nevertheless, the question must be confronted, because the answer
is in doubt. In our own time, there is a powerful current of popular
thought which is not merely unhistorical but actively antihistorical as
well. Novelists and playwrights, natural scientists and social scientists,
poets, prophets, pundits, and philosophers of many persuasions have
manifested an intense hostility to historical thought. Many of our
contemporaries are extraordinarily reluctant to acknowledge the reality
of past time and prior events, and stubbornly resistant to all arguments
for the possibility or utility of historical knowledge.

The doctrine of historical relativism was no sooner developed by
historians than it was seized by their critics and proclaimed to the
world as proof that history-as-actuality is a contradiction in terms, and
that history-as-record is a dangerous delusion which is, at best, an irrel-
evance to the predicament of modern man, and at worst a serious
menace to his freedom and even to his humanity. A few of these people
even believe, with Paul Valery, that

History is the most dangerous product which the chemistry of the mind has
concocted. Its properties are well known. It produces dreams and drunken-
ness. It fills people with false memories, exaggerates their reactions, ex-

308

CONCLUSION

acerbates old grievances, torments them in their repose, and encourages
either a delirium of grandeur or a delusion of persecution. It makes whole
nations bitter, arrogant, insufferable, and vainglorious. 1

These prejudices have become a major theme of modern literature.
Many a fictional protagonist has struggled frantically through six
hundred pages to free himself from the past, searching for a sanctuary
in what Sartre called "a moment of eternity," and often finding it in a
sexual embrace. 2

In Aldous Huxley's After Many a Summer Dies the Swan, Mr.
Propter is made to say, "After all, history isn't the real thing. Past time
is only evil at a distance; and of course, the study of past time is itself a
process in time. Cataloguing bits of fossil evil can never be more than an
ersatz for eternity." 3 In the same author's The Genius and the Goddess,
John Rivers compares history to a "dangerous drug" and dismisses it
as a productive discipline of knowledge:

God isn't the son of memory: He's the son of Immediate Experience. You
can't worship a spirit in spirit, unless you do it now. Wallowing in the past
may be good literature. As wisdom, it's hopeless. Time Regained is Paradise
Lost, and Time Lost is Paradise Regained. Let the dead bury their dead.
If you want to live at every moment as it presents itself, you've got to die at
every other moment. That's the most important thing I learned. 4

Some entertaining errors of the same sort appear in John Barth's
splendid picaresque novel, The Sot-Weed Factor, where, in sixty-five
chapters, Clio is ravished as regularly as most of the major characters.
In an epilogue, the author writes,

Lest it be objected by a certain stodgy variety of squint-minded antiquarians
that he has in this lengthy history played more fast and loose with Clio, the
chronicler's muse, than ever Captain John Smith dared, the Author here
posits in advance, by way of surety, three blue-chip replies arranged in order
of decreasing relevancy. In the first place be it remembered, as Burlingame
himself observed, that we all invent our pasts, more or less, as we go along,
at the dictates of Whim and Interest. . . . Moreover, this Clio was already a
scarred and crafty trollop when the Author found her; it wants a nice-honed
casuist, with her sort, to separate seducer from the seduced. But if, despite
all, he is convicted at the Public Bar of having forced what slender virtue the
strumpet may make claim to, then the Author joins with pleasure the most
engaging company imaginable, his fellow fornicators, whose ranks include

1. Paul Valiry, Regards sur le Monde Actuel (Paris, 1949), p. 43.

2. Jean Paul Sartre, The Reprieve (New York, 1947), p. 352.

3. Aldous Huxley, After Many a Summer Dies the Swan, Harper & Row ed. (New
York, 1965), p. 81.

4. Aldous Huxley, The Genius and the Goddess, Bantam Books ed. (New York, 1956),
p. 4.

CONCLUSION

309

the noblest in poetry, prose and politics; condemnation at such a bar, in
short, on such a charge, does honor to artist and artifact alike. 9

Other literati have set their sights on historians, rather than history.
Virginia Woolf asserted, "It is always a misfortune to have to call in
the services of any historian. A writer should give us direct certainty;
explanations are so much water poured with the wine. As it is, we can
only feel that these counsels are addressed to ladies in hoops and gentle-
men in wigs — a vanished audience which has learnt its lesson and gone
its way and the preacher with it. We can only smile and admire the
clothes." 6 Similar sentiments are cast as characterizations of historians
in Sartre's Nausea, Kingsley Amis's Lucky Jim, George Orwell's 1984,
Aldous Huxley's Antic Hay, Wyndham Lewis's Self-Condemned, Ana-
tole France's Le Crime de Silvestre Bonnard, Edward Albee's Who's
Afraid of Virginia Woolf?, Stanley Elkin's Boswell, and Angus Wilson's
Anglo-Saxon Attitudes. "It's so seldom that Clio can aid the other
muses," says one character in the latter work. "Bloody fools, these his-
torians," growls another. 7

The antihistorical arguments of our own time have infected his-
torians themselves, with serious results. Historical scholarship today
is dominated by a generation (born, let us say, between 1900 and 1940)
which has lost confidence in its own calling, lost touch with the world in
which it lives, and lost the sense of its own discipline. Historians have
failed to justify their work to others, partly because they have not even
been able to justify it to themselves. Instead, when academic historians
explain why they do history, there is a narrow parochialism and petty
selfishness of purpose which surpasses rational belief. I have heard
five different apologies for history from academic colleagues — five justifi-
cations which are functional in the sense that they permit a historian to
preserve some rudimentary sense of historicity, but only at the cost of
all ideas of utility.

First, there are those who claim that history is worth writing and
teaching because, in the words of one scholar, "It is such fun!" 8 But this
contemptible argument, which passes for wisdom in some professional
quarters, is scarcely sufficient to satisfy a student who is struggling to
master strange masses of facts and interpretations which are suddenly
dumped on him in History I. It is unlikely to gratify a graduate student,

5. John Barth, The Sot-Weed Factor, Grosset and Dunlap ed. (New York, 1964)

p. 793.

6. Virginia Woolf, "Addison," Essays, 4 vols. (London, 1966), 1:87.

7. Angus Wilson, Anglo-Saxon Attitudes (London, 1956), pp. 11, 364.

8. Fritz Stern, ed., Varieties of History: From Voltaire to the Present (New York, 1956),
p. 30.

310

CONCLUSION

who discovers in the toil and loneliness of his apprenticeship the indis-
pensable importance of a quality which the Germans graphically call
Sitzfleisch. It will not be persuasive to a social scientist who is pondering
the pros and cons of a distant journey to dusty archives. It cannot carry
weight with a general reader, who is plodding manfully through a
pedantic monograph which his conscience tells him he really ought to
finish. Nor will it reach a public servant who is faced with the problem
of distributing the pathetically limited pecuniary resources which are
presently available for social research. And I doubt that it has even
persuaded those historiographical hedonists who invoke it in defense
of their profession.

For most rational individuals, the joys of history are tempered by the
heavy labor which research and writing necessarily entail, and by the pain
and suffering which suffuses so much of our past. Psychologists have
demonstrated that pleasure comes to different people in different ways,
including some which are utterly loathesome to the majority of mankind.
If the doing of history is to be defended by the fact that some historians
are happy in their work, then its mass appeal is likely to be as broad
as flagellation. In all seriousness, there is something obscene in an argu-
ment which justifies the pedagogic torture inflicted upon millions of
helpless children, year after year, on the ground that it is jolly good fun
for the torturer.

Another common way in which historians justify historical scholar-
ship is comparable to the way in which a mountain-climbing fanatic ex-
plained his obsession with Everest — "because it is there." By this line of
thinking, history-as-actuality becomes a Himalayan mass of masterless
crags and peaks, and the historian is a dauntless discoverer, who has no
transcendent purpose beyond the triumphant act itself. If the object is re-
mote from the dismal routine of daily affairs, if the air is thin and the
slopes are slippery, if the mountain is inhabited merely by an abominable
snowman or two, then all the better! If the explorer deliberately chooses
the most difficult route to his destination, if he decides to advance by
walking on his hands, or by crawling on his belly, then better still! By this
convenient theory, remoteness is a kind of relevance, and the degree of
difficulty is itself a defense.

This way of thinking is a tribute to the tenacity of man's will but not
to the power of his intellect. If a task is worth doing merely because it
is difficult, then one might wish with Dr. Johnson that it were impossible.
And if historical inquiry is merely to be a moral equivalent to mountain-
eering for the diversion of chairborne adventurers, then historiography
itself becomes merely a hobbyhorse for the amusement of overeducated
unemployables.

CONCLUSION

311

A third common justification for history is the argument that there
are certain discrete facts which every educated person needs to know. This
view has been explicitly invoked to defend the teaching of required his-
tory courses to college freshmen, and to defend much research as well.
But it is taxonomic in its idea of facts and tautological in its conception
of education. What it calls facts are merely the conventional categories
of historians' thought which are reified into history itself. And what
it calls education is merely the mindless mastery of facts — a notion not
far removed from the rote learning which has always flourished in the
educational underworld but which no serious educational thinker has ever
countenanced.

There are no facts which everyone needs to know — not even
facts of the first historiographical magnitude. What real difference can
knowledge of the fact of the fall of Babylon or Byzantium make in the
daily life of anyone except a professional historian? Facts, discrete facts,
will not in themselves make a man happy or wealthy or wise. They will
not help him to deal intelligently with any modern problem which
he faces, as man or citizen. Facts of this sort, taught in this way,
are merely empty emblems of erudition which certify that certain formal
pedagogical requirements have been duly met. If this method is mistaken
for the marrow of education, serious damage can result.

Fourth, it is sometimes suggested that history is worth doing because
it is "an outlet for the creative urge." 9 Undoubtedly, it is such a thing. But
there are many outlets for creativity. Few are thought sufficient to justify
the employment of thousands of highly specialized individuals at a
considerable expense to society.

Tombstone rubbing is a creative act. So is the telling of tall stories.
If history is to be justified on grounds of its creative aspect, then it must
be shown to be a constructive, good, useful, or beautiful creative act. Most
people who use this argument seem to be thinking in aesthetic terms. But
if aesthetic principles become a justification for history, then surely 99
percent of the monographs which have appeared in the past generation
are utterly unjustified. Most historians publish a single book in their life-
time — usually their doctoral dissertation. I cannot remember even one of
these works which can be seriously regarded as a beautiful creative act.
There have been a good many manifestoes for creative history in the past
several decades, and more than a few essays which fulsomely describe
the potential of history as art. But the number of modem histories which
are worth reading on any imaginable aesthetic standard can be reckoned

9. Norman Cantor and Richard I. Schneider, How to Study History (New York, 1967),
p. 3. For a more extended argument, see Emery Neff, The Poetry of History (New York,
1947).

312

CONCLUSION

on the fingers of one hand. Painful as the fact may be, historians must
face up to it — literary history as a living art form is about played out.
In an earlier generation, it was otherwise. But today this tradition is either
altogether dead or sleeping soundly. An awakening has been confidently
predicted from time to time, but with every passing decade the anticipated
date has been postponed. Historians, for the past several generations,
have been moving squarely in the opposite direction. There is nothing to
suggest a change, and there are a good many hints of continuity in years
to come. Until there is a reversal, or some sort of revival, or even a single
serious and successful creative act, history as it actually is today, and as
it is becoming, must be justified by another argument.

A fifth justification for history is cast in terms of the promise
of future utility. I have heard historians suggest that their random in-
vestigations are a kind of pure research, which somebody, someday, will
convert to constructive use, though they have no idea who, when, how,
or why. The important thing, they insist, is not to be distracted by the
dangerous principle of utility but to get on with the job. It is thought
sufficient for an authority on Anglo-Saxon England to publish "important
conclusions that all Anglo-Saxonists will have to consider." 10 If enough
historians write enough histories, then something — the great thing itself
— is sure to turn up. In the meantime we are asked to cultivate patience,
humility, and pure research.

This argument calls to mind the monkeys who were set to typing the
works of Shakespeare in the British Museum. So vast is the field of past
events, and so various are the possible methods and interpretations, that
the probability is exceedingly small that any single project will prove
useful to some great social engineer in the future. And the probability
that a series of random researches will become a coherent science of
history is still smaller.

A comparable problem was studied by John Venn, some years ago.
He calculated the probability of drawing the text of Paradise Lost letter
by letter from a bag containing all twenty-six signs of the alphabet —
each letter to be replaced after it is drawn, and the bag thoroughly shaken.
Assuming that there were 350,000 letters in the poem, Venn figured the
odds at 1 in 26 350 000 , which if it were written out, would be half again as
long as the poem itself.

This operation is in some ways analogous to the method of histo-
rians who hope to construct a science of history by reaching into the grab
bag of past events and hauling out one random project after another. The
analogy is not exact — the probability of success in history is even more

10. The American Historical Review 71 (1966): 529.

CONCLUSION

313

remote than Venn's. If A is the number of possible methods (a large
number), B is the number of possible topics (even larger), C is the
number of possible interpretations (larger still), and D is the length of
a sufficient series, then the odds are 1 in (ABC) D . Now D may be as small
as 1, but A, B, or C may equal infinity. If any one of them does, then
the odds are infinitely improbable, in the sense of an infinite regression
toward zero. In this context, infinite improbability will serve as a working
definition of practical impossibility.

A series of researches can be expected to yield a coherent result only
if they are not random. If a historian hopes that his work will promote
some future purpose, then he must have some idea of what that purpose
might be. The question cannot be postponed to another day. It must be
faced now. And yet historians who justify their work as "pure research"
deliberately avoid it. Their lives are wasted in aimless wanderings, like
those which Bertrand Russell remembers from his childhood. "In soli-
tude," he writes, "I used to wander about the garden, alternately col-
lecting birds' eggs and meditating on the flight of time." 11 When grown
men carry on in this way, the results are not amusing but pathetic.

All five of these justifications for history are functional to historical
scholarship, but only in the sense that they serve to sustain a rough and
rudimentary historicity in the work of scholars who have lost their con-
ceptual bearings. But these attitudes are seriously dysfunctional in two
other ways. First, they operate at the expense of all sound ideas of social
utility. Secondly, they stand in the way of a refinement of historicity,
beyond the crude level of contemporary practice.

Academic historians have been coming in for a good deal of abuse
lately, and with a great deal of justification. There is a rising chorus of
criticism which is directed principally against the sterility and social
irrelevance of their scholarship. Only a few professional pollyannas would
assert that these complaints are without cause.

But the reform proposals that accompany these protests are worse
than the deficiencies they are designed to correct. Historians of many
ideological persuasions are increasingly outspoken in their determination
to reform historical scholarship, and often exceedingly bitter about the
willful blindness of an alleged academic establishment which supposedly
stands in their way. But these reformers are running to an opposite
error.

Historians are increasingly urged to produce scholarship of a kind
which amounts to propaganda. There is, of course, nothing new in this

11. Bertrand Russell, Autobiography, 1872-1914 (Boston, 1967), p. 14.

314

CONCLUSION

idea. It appeared full-blown in the work of James Harvey Robinson and
other so-called New Historians more than fifty years ago. 12 There was
much of it after the Second World War, in the manifestoes of conserva-
tive anti-Communist scholars such as Conyers Read, 13 and in the mono-
graphs of liberal activists during the 1950s. There is still a great deal of
it today in Eastern Europe, where more than a few historians imagine
that they are "scholar-fighters," in the service of world socialism. Today,
in America and Western Europe, this idea is being adopted with increas-
ing fervor by young radical historians, who regard all aspirations to
objectivity as a sham and a humbug, and stubbornly insist that the real
question is not whether historians can be objective, but which cause they
will be subjective to.

These scholars 14 are in quest of something which they call a "usable
past." But the result is neither usable nor past. It ends merely in polemical
pedantry, which is equally unreadable and inaccurate.

There have always been many historians who were more concerned
that truth should be on their side than that they should be on the side
of truth. This attitude is no monopoly of any sect or generation. But
wherever it appears in historical scholarship, it is hateful in its substance
and horrible in its results. To make historiography into a vehicle for
propaganda is simply to destroy it. The problem of the utility of history
is not solved but subverted, for what is produced by this method is not
history at all. The fact that earlier generations and other ideological
groups have committed the same wrong does not convert it into a right.

Moreover, the "usable" history which is presently being produced
by historians of the "New Left" is not objectionable because it is sub-
stantively radical but rather because it is methodologically reactionary.
Radical historians today, with few exceptions, write a very old-fashioned
kind of history. They are not really radical historians. A good many new
procedural devices are presently in process of development — devices
which may permit a closer approximation to the ideal of objectivity. But
one rarely sees them in radical historiography, which is impressionistic,
technically unsophisticated, and conceptually unoriginal — old concep-
tions are merely adjusted in minor respects.

If history is worth doing today, then it must not be understood
either in terms of historicity without utility, or of utility without historicity.

12. James Harvey Robinson, The New History (New York, 1912).

13. See above, p. 86.

14. For a discussion of their work, see Irwin Unger, "The 'New Left' and American
History," The American Historical Review 72 (1967): 1237-63. For a sample, see
Barton J. Bernstein, Towards a New Past: Dissenting Essays in American History (New
York, 1968).

CONCLUSION

315

Instead, both qualities must be combined. The trouble with professional
historians is that they are not professional enough — and not historians
enough. If they are to be useful as historians, then they must do so by the
refinement of their professional discipline and not by its dilution.

History can be useful, as history, in several substantive ways. It can
serve to clarify contexts in which contemporary problems exist — not by
a presentist method of projecting our own ideas into the past but rather
as a genuinely empirical discipline, which is conducted with as much
objectivity and historicity as is humanly possible. Consider one quick and
obvious example — the problem of Negro-white relations in America. It
is surely self-evident that this subject cannot be intelligently compre-
hended without an extended sense of how it has developed through time.
Negro Americans carry their history on their backs, and they are bent
and twisted and even crippled by its weight. The same is true, but less
apparent, of white Americans, too. And precisely the same thing applies
to every major problem which the world faces today. Historians can
help to solve them, but only if they go about their business in a better
way — only if they become more historical, more empirical, and more
centrally committed to the logic of a problem-solving discipline.

Historical inquiry can also be useful not merely for what it con-
tributes to present understanding but also for what it suggests about the
future. A quasi-historical method is increasingly used, in many dis-
ciplines, for the purpose of forecasting — for establishing trends and direc-
tions and prospects. Historians themselves have had nothing to do with
such efforts, which many of them would probably put in a class with
phrenology. Maybe they should bear a hand, for they have acquired by
long experience a kind of tacit temporal sophistication which other dis-
ciplines conspicuously lack — a sophistication which is specially theirs
to contribute.

Third, history can be useful in the refinement of theoretical knowl-
edge, of an "if, then" sort. Econometric historians have already seized
upon this possibility, and political historians are not far behind. What,
for example, are the historical conditions in which social stability, social
freedom, and social equality have tended to be maximally coexistent?
No question is more urgent today, when tyranny, inequality, and in-
stability are not merely disagreeable but dangerous to humanity itself.
This is work which a few historians are beginning to do. Maybe it is
time that more of them addressed such problems, more directly.

Fourth, historical scholarship can usefully serve to help us find out
who we are. It helps people to learn something of themselves, perhaps
in the way that a psychoanalyst seeks to help a patient. Nothing could
be more productive of sanity and reason in this irrational world. Histo-

316

CONCLUSION

nans, in the same way, can also help people to learn about other selves.
And nothing is more necessary to the peace of the world. Let us have no
romantic humbug about brotherhood and humanity. What is at stake
is not goodness but survival. Men must learn to live in peace with other
men if they are to live at all. The difficulties which humanity has experi-
enced in this respect flow partly from failures of intellect and understand-
ing. Historical knowledge may help as a remedy — not a panacea, but a
partial remedy. And if this is to happen, professional historians must hold
something more than a private conversation with themselves. They must
reach millions of men, and they will never do so through monographs,
lectures, and learned journals. I doubt that they can hope to accomplish
this object by literary history or by the present forms of popular history.
Instead, they must begin to exploit the most effective media of mass
communication — television, radio, motion pictures, newspapers, etc.
They cannot assign this task to middlemen. If the message is left to com-
munications specialists, it is sure to be garbled in transmission. All of
these uses of history, as history, require the development of new strategies,
new skills, and new scholarly projects.

In addition to these four substantive services which historians can
hope to provide, there is another one which I regard as even more im-
portant. Historians have a heavy responsibility not merely to teach people
substantive historical truths but also to teach them how to think histori-
cally. There is no limit to the number of ways in which normative human
thinking is historical. Nobody thinks historically all the time. But every-
body thinks historically much of the time. Each day, every rational being
on this planet asks questions about things that actually happened —
questions which directly involve the logic of inquiry, explanation, and
argument which is discussed in this book.

These operations rarely involve the specific substantive issues that
now engage the professional thoughts of most historians. They do not
touch upon the cause of the First World War, or the anatomy of revolu-
tions, or the motives of Louis XIV, or the events of the industrial revolu-
tion. Instead, this common everyday form of historical thought consists
of specific inquiries into small events, for particular present and future
purposes to which all the academic monographs in the world are utterly
irrelevant.

Historical thought ordinarily happens in a thousand humble forms —
when a newspaper writer reports an event and a newspaper reader peruses
it; when a jury weighs a fact in dispute, and a judge looks for a likely
precedent; when a diplomat compiles an aide-memoire and a doctor con-
structs a case history; when a soldier analyzes the last campaign, and a
statesman examines the record; when a householder tries to remember

CONCLUSION

317

if he paid the rent, and when a house builder studies the trend of the
market. Historical thinking happens even to sociologists, economists, and
political scientists in nearly all of their major projects. Each of these
operations is in some respects (not all respects) historical. If historians
have something to learn from other disciplines, they have something
to teach as well.

The vital purpose of refining and extending a logic of historical
thought is not merely some pristine goal of scholarly perfection. It in-
volves the issue of survival. Let us make no mistake about priorities. If
men continue to make the historical error of conceptualizing the problems
of a nuclear world in prenuclear terms, there will not be a postnuclear
world. If people persist in the historical error of applying yesterday's
programs to today's problems, we may suddenly run short of tomorrow's
possibilities. If we continue to pursue the ideological objectives of the
nineteenth century in the middle of the twentieth, the prospects for a
twenty-first are increasingly dim.

These failures — failures of historical understanding — exist every-
where today. Frenchmen, in pursuit of their venerable vision of Gallic
grandeur, combine a force de frappe with the fallacy of anachronism — a
lethal combination. Arabs cry up a jihad against the infidels, as if nothing
had changed in nine hundred years but the name of the enemy. On the
other side of the Jordan River, Jews nurse their bitter heritage of blood
and tears, without any apparent sense of how the world has changed. In
Moscow and in Washington, in London and in Bonn, in Peking and New
Delhi, statesmen and citizens alike are unable to adjust their thoughts
to the accelerating rate of changing realities.

That people will learn to see things as they are — that they will
understand the world as it is, and is becoming — that they will become
more rational and empirical in their private thoughts and public policies
— that these things will come to pass, is not what Damon Runyon would
have called a betting proposition. He might have figured the most
favorable odds at six to five, against. But if people continue to commit
their fatal fallacies at something like the present rate, the odds for their
survival will become a long shot.

Responsible and informed observers have estimated that by the
1990s as many as forty-eight nations may possess nuclear weapons. 15 As
the number of these arsenals increases arithmetically, the probability
of their use grows in geometric ratio. Biological and chemical weapons
of equal destructive power and even greater horror are already within the
reach of most sovereign powers, and many private groups as well.

15. Sir John Cockcroft, "The Perils of Nuclear Proliferation," in Nigel Calder, ed.,
Unless Peace Comes (New York, 1968), p. 37.

318

CONCLUSION

Natural scientists have helped to create this deadly peril; now it is
the business of social scientists to keep it in bounds. Here is work for
historians to do— work that is largely educational in nature — work that
consists in teaching men somehow to think reasonably about their con-
dition. Reason is indeed a pathetically frail weapon in the face of such
a threat. But it is the only weapon we have. To the task of its refine-
ment, this book has been addressed.

INDEX

(Works discussed in this book are not indexed by title. Consult the name of the author.)

Aaron, Daniel, 239
Abduction, see Logic, abductive
Abel, Reuben, 12
Abernethy, Thomas P., 31
Abolitionists, 10, 32-3, 117, 273-4, 283
Absolute priority, fallacy of, 178
Absolutism, history of, 11
Abundance, history of, 172
Academic freedom, 133
Accent, fallacy of, 271-5
Accident, fallacy of, 306
Actium, battle of, 174
Acton, Lord, 7n, 194n, 281, 307
Ad antiquitam, fallacy of argument, 2S7-8
Ad baculum, fallacy of argument, 254-6
Ad consequentiam, fallacy of argument,
300-2

Ad crumenam, fallacy of argument, 293-4
Ad hominem, fallacy of argument, 91,
290-3; abusive, 290; circumstantial,
291; associative, 291; tu quoque, 292
Ad invidiam, fallacy of argument, 304
Ad metum, fallacy of argument, 304
Ad misericordiam, fallacy of argument, 304
Ad modum, fallacy of argument, 79, 304
Ad nauseam, fallacy of argument, 302-3
Ad novitum, fallacy of argument, 71,
299-300

Ad odium, fallacy of argument, 304
Ad populam, fallacy of argument, 51-3,
294, 302

Ad superbiam, fallacy of argument, 304
Ad superstitionem, fallacy of argument,
304

Ad temperantiam, fallacy of argument,
296-7

Ad verecundiam, fallacy of argument,

283-90
Adams, Brooks, 139
Adams, Henry B., 139, 268, 298, 300
Adams, Herbert Baxter, 179-80
Adams, John, 56, 109-10, 172, 210-11,

273

Adams, Samuel, 105
Addled Parliament, 245
Adduction, see Logic, adductive
Administrative history, 142
Aesthetic fallacy, 87-90, 99

Aesthetics, and history, 87-90, 137,
155-6, 167, 254, 265, 268-70,
311-12

Africa, history of, 227; see also
specific countries and colonies

Agriculture, history of, 118, 142

Alamo, battle of the, 72

Albee, Edward, 309

Alcohol, and road accidents, 169

Alden, John R, 267n

Alexander I, 138

Alexander of Greece, 174

Alienation, history of, 218

Ambiguity, fallacy of, 11, 35, 145,
265-67

America, history of, see specific entries
American Civil Liberties Union, 134
American Historical Association, 43,

85-6, 94, 211
American Jewish Historical Society, 141,

144

Ames, Nathaniel, 273
Amis, Kingsley, 309
Amory, Cleveland, 227
Amphiboly, fallacy of, 267-8
Anachronism, fallacy of, 11, 49, 132-5,

157-9, 199, 242, 250, 317
Analogy, 100, 243-59
Analytical definition, 279
Analytical philosophy of history, see

Epistemology
Ancestors, argumentative appeals to,

298-9

Ancient history, 6, 141; see also Egypt,

Greece, Rome, etc.
Anderson, F. H, 4n
Anderson, Maxwell, 54
Andrea Doria, 166
Andrew, John, 20
Andrews, Kenneth R., 286n
Angle, Paul M., 29 In
Anglicanism, 222
Annates, 218

Anthropology and history, 81, 229; see

also Culture
Anthropomorphism, 190
Anthropopathism, 190
Anticommunism, and history, 86-94

320

INDEX

Antietam, battle of, 173
Antifederalists, 77, 84, 105-6
Antihistory, 307-9
Antinomian fallacy, 94-7, 99
Antiphrasis, 269
Antiquarian fallacy, 140-2
Antiquitam, argument to, 298-9
Antiracism, fallacy of, 233-6
Anti-Whig history, 139
Apathetic fallacy, 193-5
Apel, Willi, 124n
Apophasis, 269
Aposiopesis, 269

Appositive proof, fallacy of, 56-8
Aptheker, Herbert, 274-5
Arabs, history of, 74, 147, 185, 317
Aravisci, history of, 138
Archaeology, and history, 41
Archetypes, fallacy of, 150, 192
Argument, and history, xv, 261-306
Aries, Philippe, 146n
Aristocracy, history of, 22-4
Aristotelianism, 245
Aristotle, 87, 164, 263, 306
Armada, see Spanish Armada
Armstrong, Louis, xii
Aron, Raymond, 42n, 95
Art history, 137-8, 142; see also

Aesthetics
Arthurian legend, 152, 198
Ashkenazic Jews, 233
Ashley, Margaret, 6n
Asia, history of, 246; see also specific

countries and colonies
As-if questions, see Fictional questions
Asquith, Herbert, 25
Assassinations, 1 68
Athens, history of, 124
Attribution, misuse of, 287
Auerbach, Jerold S., 53-5
Augustan age, 245
Aulard, Alphonse, 292-3
Austin, Benjamin, 106
Austria, history of, 192
Austro-Hungarian Empire, 2 1
Authority, history of, 14, 218; misuse

of, see Ad verecundiam
Avant garde, 245
Averaging, 118
Axis, 245

Aydelotte, W. O., 92, 114n

Ba Maw, 228n
Babylonian captivity, 245
Bacon, Francis, 4n, 6, 221
Baconian fallacy, 4-8
Baculum, argument ad, 294-6
Bagehot, Walter, 160
Bailyn, Bernard, xiin, 34, 93n
Bainville, Jacques, 304
Bajazet, 174

Baldwin, Stanley, x-xi
Balliol College, 33
Bancroft, George, 153, 226
Baptist churches, 48
Barbary Wars, 227
Barbed wire, history of, 141
Barber, Elinor, 218n
Barnburners, 245

Barraclough, Geoffrey, 138, 141n

Barrow, Thomas, 14-5

Barruel, Abbe, 304

Barth, John, 77, 308-9

Barzun, Jacques, 58, 67, 151n, 231

Beach, Moses, 1 1 6

Beamish, T., 145n

Beard, Charles, 26, 27, 29-30, 41,

42n, 76-7, 105, 143, 164, 165n, 181n,

296-7, 299
Becker, Carl, x, 40, 41, 57, 221, 270
Beckley, John, 140
Beer, Samuel, 218n
Beerbohm, Max, 25
Behaviorism, 91, 209
Belief units, 217
Bemis, Samuel Flagg, 302
Benson, Lee, 13n, 37, 92, 113-16,

168, 276-7
Bentham, Jsremy, xviin, 221, 274,

282-3, 297, 304, 306
Benthamites, 245
Bercovitch, Sacvan, 199n
Berenson, Bernard, 1 37-8
Berkeley, George, 128n
Berkeley, University of California at,

27

Bernstein, Barton J., 148n, 314n
Best, John H., 133n
Bill of Rights, 133
Bills of Attainder, 133
Binion, Rudolph, 188n, 194n
Biochemistry and history, 233
Biography and history, 193
Bismarck, Otto von, 138
Bjork, Gordon C, 127
Black Caribs, 234
Black, Cyril E., 238-9
Black Legend, in Latin American history,
31

Blackmore, Thaung, 228n
Black or White fallacy, 276-7
Blacks, see Negroes
Bladensburg, battle of, 141
Blaine, James G., 168
Blake, Christopher, 43n, 18 In
Blake, Robert, 132n
Blanc, Louis, 304
Blank check, 245

Bloch, Marc, 145, 180, 218n, 286,

290, 299
Bloody Assizes, 245

INDEX

321

Boas, George, 24
Bober, M. M., 194n
Bodleian Library, 67
Bohme, Helmut, 218n
Bolingbroke, Henry, 225n, 284
Bond, Horace M., 26n
Boorstin, Daniel, 57, 96, 155, 156
Booth, John Wilkes, 16
Bogue, Allan G., 114n, 218n
Boston Athenaeum, 139, 140
Boswell, James, 225n, 292
Boulding, Kenneth E., 68, 238n
Bowen, Catherine Drinker, 89n
Bower, Robert T., 107n
Bowie, Jim, 72
Bowles, E., 91 n
Boxers, 245
Boyd, Julian, 299
Bradley, A. C, 288
Braudel, Fernand, 218n
Braun, Werner von, 257
Bridenbaugh, Carl, 43-5, 94, 121
Bridgman, Percy W., 38n, 264
Brighton, 40
Brinkmanship, 245
Brinton, Crane, 34
Brockunier, S. H., 133n
Brodie, Fawn, 188n
Bromsen, Maury A., 232n
Brooke, Lord, 23
Browder, Earl, 291, 299
Brown, B. Katherine, 22-3
Brown, Norman O., 195n
Brown, Robert E., 107-8, 296
Brown, Theodore A., 285n
Brumfitt, J. H, 240n
Bruner, J., 157n
Brunton, D., 95n
Bryan, George, 105
Bryan, Samuel, 105, 106
Bryan, William Jennings, 1 10
Bryce, James, 243
Bubble Act, 245

Buchez, Pierre Joseph Benjamin, 304

Bukharov, B. I., 108

Buckley, William, 195

Bunge, Mark), 164

Burckhardt, Jacob, 31, 82

Burden of proof, 63

Burma, history of, 228-9

Bury, J. B., 87n, 179

Bushinski, E. A., 256n

Butler, Nicholas Murray, 76

Butterfield, Herbert, 139, 281

Byron, Lord, 16

Cabals, 217

Cady, John F., 228, 229n
Cairo, 159

Cajetan, Cardinal, 255n

Calder, Nigel, 317n

Calhoun, John, 110

Calvin, John, 270

Calvinism, 78, 90

Camarillas, 217

Cameralism, 245

Campbell, Lilly B., 81 n

Cantor, Norman, 31 In

Capitalism, history of, 168, 178,

245, 265, 275
Caravans, 217
Carbonari, 245
Caribs, 234

Carlyle, Thomas, x, 91, 207, 304
Carnap, Rudolf, xv
Carpetbaggers, 182
Carr, E. H., 13, 37, 41, 103, 139n,

157, 187n, 207, 249, 281
Carroll, Lewis, 284, 302
Carter, Hodding, 72n
Cash, W. J., 108, 219-221
Caskey, Willie M., 27n
Cason, Clarence, 108
Cassirer, Ernst, 3, 82n, 221
Caste systems, 217, 223
Catachresis, 269
Catherine II, 289
Catton, Bruce, 72n
Caucuses, 217

Causation, xv, 13, 34, 100, 163,

164-186
Censorship, 133
Cervantes, 167

Chambers, William Nisbet, 139-40,
218n

Change, presumptive, fallacy of, 154
Change, see Narration
Channing, Edward, 3 1
Chapman, Gerald W., 195-6
Character, national, see National

character
Charles II, 8
Charles VI, 61

Charleston, S.C., history of, 45

Cherokee Indians, 10

Chesterton, G. K., 191-2

Chicago, 111., 107

Childe, Gordon, HOn

Childhood, history of, 49

China, history of, 22, 137, 147, 151,

177, 218, 246
Chou En-lai, 22
Christian, John L., 228
Christianity, history of, 147, 152,

179, 185, 279
Chronic fallacy, 152-3
Chronicles, and history, 152-3,

160-1

Chronology, 41, 152-3; see also Narration

Church history, 217

Churchill, Randolph S., xin, 159n

322

INDEX

Churchill, Winston, xin, 97, 131,

159, 174, 246
Cicero, 88
Ciceronian, 245
Cigarettes, and cancer, 169
Cincinnati, Ohio, 46
Circular proof, fallacy of, 49-51,

209

Citations, misuse of, 286

Cities, history of, see Urban history

Civilization, 81, 216, 270

Civil liberties, history of, 55, 132-3

Civil War, American, 13, 32, 72, 173,
179, 186, 212-13

Civil War, English, 201; see also Puri-
tans

Clans, 217

Clapham, Sir John, 113

Clark, Sir George, 73

Clark, Manning, 287

Class, social, in history, 84, 217,

265; see also Groups
Cleopatra, 174
Cleveland, Ohio, 46-7
Cleveland, Grover, 46
Clinton, George, 105, 106, 125-6
Cliometrics, see Economic history
Cliques, 217
Clubs, 217
Cobb, Frank, 54-5
Cobban, Alfred, 37n, 221
Cochran, Thomas C, I46n
Cockcroft, Sir John, 3l7n
Cohen, J. B., 1 18n

Cohen, Morris Ralph, 128n, 155n, I78n
Cold War, 85, 245
Collective bargaining, 1 1
Collectives, 217
Colleges, 217

Collingwood, R. G„ xii, 3, 39, 183-5,

196-7, 198
Collis, Maurice, 228

Colonial history, American, 22-4, 43, 142;

see also Puritans, Virginia, etc.
Columbia University, 76, 135
Combinations, 217

Commager, Henry Steele, 143, 190-1,

239, 299
Committees, 217
Communism, see Radicalism
Community, 216
Comparative history, 57-8, 238
Comparative Studies in Society and

History, 218
Composition, analysis of, 216-42
Composition, fallacy of, 219-21
Computers and history, 68, 90; see also

Quantification
Comte, Auguste, 66
Concentration camps, 256

Conceptualization, 160

Conditional questions, see Fictional

questions
Condorcet, Antoine, 57, 66, 221
Confederacy, Southern, history of, 173
Confucius, 22

Congregationalist churches, 48, 121
Congregations, 217
Congress, U.S., 78, 111, 116, 217
Conjecture, xix; see also Creativity
Conner, R. D. W., 27n
Connections, 217
Connotation, 274
Consensus history, 20, 155
Consequentiam, argument ad, 300-2
Conservatism, and history, 69, 85, 156,

158-9, 190, 208, 242
Consortiums, 217
Constable, Giles, 266
Constantinople, 23 1
Constitution, Mass., 47; U.S., 29,

74-8, 83, 84, 105
Constitutional history, 142
Constructive definition, 278
Contemporary history, 138
Context, 63

Continuity, presumptive, fallacy of, 154
Contradictory questions, fallacy of,
34-6

Control groups, statistical, 117

Conventions, 217

Conway, Alan, 27

Coolidge, Mrs. Calvin, 267

Coolidge, Calvin, 31, 133

Coon, Carleton S., 232n, 233

Copernicus, 136-7, 161

Cordon sanitaire, 245

Correlation, statistical, 47-8, 113-15,

168-9; Pearson, 119; Spearman Rank

Order, 115
Corry, John, 7ln
Cosmology, 255
Coteries. 217
Cotton, John, 22-4
Cotton, price history of, 14
Counterfactual questions, see Fictional

questions

Counterquestions, fallacy of, 28-31, 77,

199, 208, 246
Covering law model, see Hempelian model
Cox, Archibald, 135n
Cox, Eugene L , 74n
Cox, La Wanda and John H., 203n
Cox Commission, 135
Crafts, 217

Craven, Wesley Frank, xn

Creativity, xix, xx, 243, 306, 311-12

Crevecoeur, J. Hector, 171

Critical period, of American history, 31

Croce, Benedetto, 42

Crocker, Lester, 221

INDEX

323

Croix de Feu, 245

Cromwell, Oliver, 83, 159, 199

Crosse, Henry, 50n

Cross grouping, fallacies of, 236-40

Crowds, 217

Crowe, Charles, 148n

Crumenam, argument ad, 293-4

Ctesias, 300

Cults, 217

Culture, definitions of, 217; history and,

67, 96, 216, 230-1, 246, 265
Cum hoc, propter hoc, fallacy of, 167-8
Cummings, E. E., 42
Cunningham, Noble, U7n, 126-7
Currency estimates, 44
Current, Richard, 239
Czechoslovakia, 248

Daalder, Hans, 218n
Dahrendorf, Ralf, 203n, 218n
Daily Worker, 108
d'Alembert, Jean, 57
Dangerfield, George, 212
Daniels, Josephus, 55
Danish colonies, 80
Dante, 230

Danto, Arthur C, xi, 42n, 64n,

66n, 131n, 162, 181n, 195n
Dark Age, 245
Darling, Edward, 53n
Darwin, Charles, 193
Dating, 133

Davidsohn, Robert, 300
Davis, Kenneth, 73n, 272n
Davis, Lance, 7, 18n
Davis, William W„ 26n
Daylight-saving time, 168
Debbins, William, 196n
Declaration of Independence, 273
Declarative questions, fallacy of, 24-8
De Coulanges, Fustel, 6, 7n
Deduction, see Logic, deductive
Deductive law model, see Hempelian
model

Deference, history of, 147

Definition, 264-5, 277-80

De Gaulle, Charles, 75, 159

De homine, fallacy of argument, 292

Delacroix, Ferdinand, 190

Demartino, Manfred F, 213n

Democratic Party, 107, 110-11, 114; see

also Jeffersonian Republican Party

and Jacksonian Party
Democratization, 22-4, 58, 179, 265,

280

Demography, and history, 120-1, 218,
219

De Morgan, Augustus, xvii, 50-1, 53n

Denominations, 217

Depressions, 245

De Sade, Marquis 219

Desai, Meghnad, 18 n
Destler, Chester McA., 46-7
Detail, misuse of, 285, 288
Determinism, x, 13, 189, 194-5; see also

Economic determinism; Materialism;

Psychic determinism
Deuel, Leo, 300n
Deutsch, Karl, 192
De Vio, Thomas, 255n
De Voto, Bernard, 89
Dialectics, 28-9

Dichotomous questions, fallacy of, 9

Dichotomy, 9

Didactic fallacy, 157-9

Diderot, Denis, 221, 289

Difference, fallacy of, 222-3; converse

fallacy of, 223-4
Dionysius of Halicarnassus, 78
Diplomatic history, 142, 218
Dispersion, statistical, 118
Disraeli, Benjamin, 108, 132
Disraeli, Isaac, 288
District of Columbia, 117
Division, fallacy of, 221-2
Dobzhansky, Theodosius, 232n
Documents, preservation of, 142
Dodds, E. R., 188n, 194
Dogberry, 269
Doherty, Robert, 279
Donaghan, Alan, 196n
Donald, David, 27, 111-12, 117, 283-4
Donleavy, J. P., 77
Dorolle, Maurice, 244n
Dos Passos, John, 73n
Dostoevsky, Feodor, 15, 108
Doubleday & Co., 71
Double-reversing generalization, fallacy

of, 125
Doubt, 53
Doughface, 245
Dowdey, Clifford, 73n
Dray, William, 128n, 196n
Dred Scott decision, 10
Dreyfus, Alfred, 118
Drouet, 16

Drugs and history, 52
Duane, James, 105, 106
Du Bois, W. E. B., 26, 27
Duff, J. D., 88n
Dumont, Etienne, 282n
Duncan, Otis Dudley, 119n
Dunne, Finley Peter, 125
Dunning, William Archibald, 25-7
Dutch colonies, 80
Dutchess County, N.Y., 84

Eastbourne, 40
Easton, David, 209
Eaton, Peggy, 72
Ecclesiastical history, 142
Eckenrode, Hamilton, 26n

324

INDEX

Ecological fallacy, 119-20
Economic determinism, 26, 29, 30, 196,
203

Economic history, 7, 16-21, 72, 94, 123,

127-8, 142, 177, 218, 246, 275
Economic man, see Economic determinism
Edinger, Lewis J., 188n
Edison, Thomas, xviii
Education, history of, 177, 217, 218, 265
Edward VII, 71
Edwards, Jonathan, 96, 199
Egypt, 212
Einstein, Albert, 130
Eisenhower, Dwight D., 20
Elections, English, of 1830, 230; U.S.,

of 1808, 123; 1812, 123; 1828, 128;

1844, 113; 1860. 168; 1884, 168;

1892, 265; 1916, 55; 1936, 106; 1952,

265; 1956, 272; 1968, 276
Eliade, Mircea, 150-2
Eliot, Charles W, 293
Eliot, T. S., 20, 155
Elitism, fallacy of, 208, 230-2
Elizabeth I, 266-7

Elkins, Stanley N., 8n, 57-8, 76n, 105-6,

205-6, 256-7, 309
Elliott, J. H., 184n
Ellipsis, 269
Ellsworth, Oliver, 77
Elsbree, Willard, 228
Elton, G. R., 7, 64, 67, 97n, 165,

296n, 301
Emancipation Proclamation, 173
Emmet, E. R., 51

Empiricism, xxi, 12, 12-13, 25, 27,

65, 68-9, 254, 255
Empson, William, 265n
Encina, Francisco A., 42n
Enculturative groups, 217
England, history, of, 20, 23, 58-9, 73-4,

81-2, 86, 93, 97, 122, 147, 150, 167,

192, 198, 229-30, 266
English colonies, 80
English reform bill, 78
Enlightenment, history of, 57, 70, 190,

204, 221, 245
Enthymemes, 181
Enumerative definition, 278, 281
Epigraphy, 41
Episcopalianism, 1 2 1
Epistemology, xi-xii
Equivocation, fallacy of, xvii, 274-5
Erikson, Erik, 49, 188n, 301
Error, xvii

Essences, fallacy of, 60, 68-70, 88, 94,

99, 126, 143, 281, 306
Essex Junto, 57
Etceteration, 266-7
Ethics, and history, 79, 155-6, 182,

187, 188; see also Moral judgments;

Moralizing; Moralistic fallacy

Ethnocentrism, fallacy of, 226-30
Ethnomorphism, fallacy of, 224-6
Euphemism, 269

Europe, history of, 35, 70, 138; see
also specific countries

Events, defined, xv; in history, 66

Eversley, D. E. C, 219n

Evidence, 62-3, 40-63

Exceptions, 127

Excluded middle, law of, 10

Explanation, denned, xv; in history, xv-
xvi, 34, 101-260; by generalization,
103-30; by narration, 131-63; by
causation, 164-86; by motivation, 187-
215; by composition, 216-42; by
analogy, 243-60

Exposition, 100

Extrapolation, 120-1, 284

Ezell, John S., 148n

Fabians, 245
Factions, 217
Factors, 165-6

Facts, 40-63, 66, 70, 109-10; denned, xv;

particularity of, 5; truth value of,

5; as arguments, 263
Factual significance, 64-100
Factum probandum, 62
Factum probans, 62
Fallacies, defined, xvii; nomenclature of,

4n, 282; cautions against misuse of,

xviii, xx, 283, 305-7; see also Index

of Fallacies
Fallacist's fallacy, 305
False analogy, fallacy of, 251-2
False dichotomous questions, fallacy of, 9
False interpolation, fallacy of, 122
False periodization, fallacy of, 145
Family, history of, 49, 217
Fanfani, Amintore, 168
Farrand, Max, 30
Farrington, Benjamin, 4n
Farwell, Byron, 274

Federalist Party, 56, 58, 105-6, 109-10,
126, 202, 209, 210, 222, 245

Federal Reserve Board, 182

Fehrenbacher, Don E., 8-9

Feis, Herbert, 159, 211

Fennell, J. L. I., 200

Ferguson, Adam, 57

Ferguson, E. James, 30n

Ferguson, Wallace K„ 146n

Fertig, James W., 26n

Feudalism, 11, 103, 180, 245, 265

Ficklen, John R., 26n

Fictional questions, fallacy of, 9,
15-21

Fictions, 19

Fiedler, Leslie, 155

Figments, 19

Figurative definition, 278

INDEX

325

Figures, fallacy of, 51, 268-71
Figures of speech, 268-70
Filibuster, 245

Fischer, David Hackett, 298n
Fischer, Fritz, 164n
Fischer, Roger A., 148n
Fisher, H. A. L., 124
Fishlow, Albert, 7, 218n
Fiske, John, 31
Fitzgerald, C. P., 177n, 246n
Fitzgerald, Scott, 135
Flaubert, Gustave, 304
Fleming, Donald, 212
Fleming, Walter L., 26n
Flemings, 34

Fogel, Robert W., 7, 16-19, 21, 218n

Fog-of-war technique, 212-13

Forecasting, 315

Forester, C. S., 72n

Formal logic, see Logic, formal

Fortuitous fallacy, 97-99

Founding fathers, 245

Fowler, H. W„ 257, 269, 270

Fox, Charles James, 246

Fox, George, 278, 281

France, history of, 145, 147, 231,

250, 317
France, Anatole, 309
Franco-Prussian War, 6, 250
Frankel, Charles, 42, 18 In
Franklin, Benjamin, 106, 133, 153,

171, 244, 259
Franklin, John Hope, 224
Free security, 170

Freedom, in historical scholarship, 27
Freeman, Douglas Southall, 134, 212,
300

French and Indian War, 171, 227
French colonies, 80

French Revolution, see Revolution, French
Freud, Sigmund, 49, 188-9, 192, 195n,
226

Freudian history, 174, 187-215, 226, 245

Friedel, Frank, 239

Friends, Society of, see Quakers

Fromm, Erich, 194n

Fronde, 245

Frontier, history of, 31, 170, 212, 294

Frost, Robert, 130

Froude, James, 300

Fuller, J. F. C, 158n

Furnivall, John S., 228

Furtive fallacy, 74-8, 210

Fussner, F. Smith, 4n, 147

Gabriel, Ralph Henry, 299
Gag rule, 245
Galileo, 244, 259
Gallie, W. B., xiin
Gamblers' fallacy, 118
Gambrell, H. P, 133n

Gamow, George, 244
Gangs, 217

Gardiner, Patrick, 104n, 128n

Garfield, James A., 168

Garland, Robert, 168

Garn, Stanley, 23 2n, 233-4

Garner, James W., 26n

Garraty, John A., 152-3

Gash, Norman, 229, 230n

Gatell, Frank O., 116n

Gaustad, Edwin S„ 121n

Gautier, Etienne, 219n

Gaxotte, Pierre, 304

Gemeinschaft, 216

General interpretations, 5

Generalization, xv, 51, 100, 184, 209;
varities of, 103-4; statistical, xv,
103-30; double-reversing, 125-6;
insidious, 124-5; hasty, 104-10,
114-116, 120-24; nonsensical, 116-18;
halting, 127

General laws, 128; see also Hempelian
model

Generations, analysis of, 120, 217
Genetic fallacy, 155-7
Genetics, 233

Genovese, Eugene D., 13n, 28
Gens, 217

Gentlemen's Agreement, 245

Genus and difference, definition by, 277-8

George III, 192, 252

George, Alexander and Juliette, 188n

Germany, history of, 11, 29, 80, 159,

171, 181, 218, 294-5
Germ theory, 179-80
Gerry, Elbridge, 105, 106
Gerschenkron, Alexander, 18n, 19n
Gesellschaft, 216
Gettysburg, battle of, 16, 132
Geyl, Pieter, 13, 34, 139n, 280n, 293n
Gibbon, Edward, 174
Gide, Andre, 42
Gilchrist, David, 127n
Gillespie, C. C. 16 In
Gipson, Lawrence Henry, 288, 297n
Glaab, Charles N„ 285n
Glass, D. V., 219n
Glazer, Nathan, 8
Goldman, Eric, 21-2, 159n, 296
Goldsmith, Oliver, 292
Goldsmith, Raymond, 123
Goldwater, Barry, 159-60, 195
Gooch, G. P., 6
Good, I. J., 39n
Goodman, Leo, 119n
Goodman, Nelson, 15, 16n
Good Neighbor Policy, 245
Gorer, Geoffrey, 192-3
Gorham, Nathaniel, 1G6
Gottschalk, Louis, 104n

326

INDEX

Goubert, Pierre, 219n

Graff, Henry, 58

Grammar, and history, 306

Grand Peur, 245

Graves, Robert, 88n

Great Awakening, 112

Great Britain, 250; see also England; Ire-
land; Scotland; Wales

Greece, history of, 87-8, 124, 147, 174,
185, 191, 194, 231, 249

Greer, Donald, 95n

Gregorian chant, 124

Greven, Philip J., Jr., 218n

Grew, Joseph C, 210

Grimke sisters, 273

Gripsholm, 166

Groups, xv, 13, 100, 216-42; defined,

216-17
Guedalla, Philip, 15
Guelphs, 245
Guilds, 217
Guizot, Francois, x
Guttmacher, Manfred, 192

Hacker, Louis M., 18n

Hadas, Moses, 232n

Haig, Douglas, 158

Halacy, D. S., 258n

Halevy, Elie, 229-30

Halicarnasus, 74

Hall, D. G. E , 228

Hallam, Henry, 281

Haller, William, 32

Hamburger, Joseph, 78

Hamilton, Alexander, 96, 105, 109-10,

153, 202, 210, 246
Hamilton, Earl J., 184n
Hamilton, Edward, 72n
Hamilton, J. G. de Roulhac, 25n,

26n

Hamilton, William Gerard, 28n, 282-3

Hamlets, 217

Hampton, Wade, 20

Handlin, Mary, 47-8

Handlin, Oscar, 47-8, 302-3

Harding, Warren G., 168

Hardy, Thomas, 145

Harrison, Joseph H., Jr., 133n

Harrison, William Henry, 168

Hart, H. L. A., 185n

Hartz, Louis, 56, 155, 208-9, 210

Harvey, G. E., 228

Hastings, battle of, 40

Hats and Caps, 245

Hawaii, 233

Hawke, David, 13 n

Hawke, G. R., 18n

Hays, Samuel, 92

Heaton, John L., 53n, 54

Heavenly kingdom, 245

Hecto-history, 145

Hegel, G. W. F., 31, 65-6, 156,

157-8, 195n
Heimert, Alan, 112
Helmholtz, Hermann, 164
Helvetius, Claude Adrien, 221
Hempcl, Carl G„ 38n, 128-30
Hempelian model, xi, 128, 181
Henry VI, 61
Henry VIII, 59
Henry, Louis, 219n
Henry, Patrick, 105, 106
Herder, Johann, 156
Herlihy, David, 266n
Hermann, Charles F. and Margaret G., 21
Herndon, William H , 291, 297
Herodotus, 71, 74, 300
Heuristic learning, 25, 27
Hexter, J. H., 142, 150, 161,

162-3, 177
Hicksites, 280
Higginson, Stephen, 268
Higham, John, 9n
Highet, Gilbert, 232n
Hildreth, Richard, 300
Hill, D. H., 173
Himmelfarb, Gertrude, 78n
Hirsch, E. D., 155n

Historian, defined, xv; as scientist, xxi-xxii

Historians' fallacy, 209-13

Historical idealism, see Idealism

Historical man, 203

Historical materialism, see Materialism

Historical relativism, see Relativism

Historical thought, logic of, see Logic

Historicism, 155-7, 188

History, defined, xv; history of, 147;

philosophy by examples, 78; uses of,

315-18
History and Theory, 218
Hitler, Adolf, 93-4, 158, 227, 236,

246, 248
Hobbes, Thomas, 199, 201-2
Ho Chi Minh, 249, 250, 299
Hoffding. Harald, 255
Hofstadter, Richard, 75, 110, 195, 239
Hogben, Lancelot, 115n
Hoggart, Richard, 85n
Holbach, Paul, 57, 221
Holbrook, Stewart, 73n
Holist fallacy, 5, 42, 65-8, 99, 145,

354-5

Holistic analogy, fallacy of, 254-5
Holland, history of, 34
Hollis, J. P., 26n
Holroyd, Michael, 98n
Hoist, Hermann von, 301
Homer, 75

Hominem, argument ad, 291-3
Honore, A. M., 185n
Hook, Sidney, xiin, 19n
Hoover, Herbert, 110, 159, 211

INDEX

327

Homey, Karen, 189
Horowitz, Henry, 145n
Housman, A. E., 40
How questions, 14-15
Howe, Laurence Lee, 67n
Hsu Cho-yun, 218n
Huff, Darrell, 118n
Huizinga, Johan, 4, 31, 6 In, 177
Humanism, 91-97
Humboldt, Alexander von, 300
Hume, David, xv, 57, 204, 221
Hundred Years, War, 209
Hunt, E. H., 18n
Hutchinson, Thomas, 23n
Huxley, Aldous, 42, 308, 309
Huygens, Christiaan, 244
Hyderabad, India, 168
Hyman, Herbert H, 226n
Hypallage, 269

Hyperbole, 52, 58, 269, 270-1
Hypostatized proof, fallacy of, 55-6
Hypothesis, xvi, 3, 4, 6, 7, 8, 19,

24-5, 36-8, 39
Hysteron proteron, 269

IBM Corporation, 90
Iceland, 225

Idealism, historical, xii, 13, 156,

195-200
Idealist fallacy, xii, 195-200
Identity, fallacy of, 167, 177
Ideological groups, 217
If questions, see Fictional questions
Iggers, Georg G., 156n
Ignoratio elenchi, fallacy of, 91, 284
Imagery, in historical writing, 51,

268-70

Immigration, history of, 302-3
Imperialism, 245
Impressionism, 104, 124-5
India, history of, 137-223
Indians, American, 225, 305-6
Indiscriminate Pluralism, 175-7
Indispensability, see Inevitability
Individuality, 13
Indonesia, history of, 218, 229
Induction, see Logic, inductive
Industrialization, history of, 11,

52, 59, 149, 245
Inevitability, in history, 12-13, 17
Inference, xvi, 215
Innovation in history, 218
Innuendo, 273-4, 281
Inquiry, ix, 1-100; question framing,

1-39; factual verification, 41-63;

factual significance, 64-100
Insidious analogy, fallacy of, 244-7
Insidious generalization, fallacy of,

124-5

Institutional history, 142

Intellectual history, 142, 190, 195-

200; see also Motivation
Interminable fallacy, 149-50
International Association of Machinists,

168-9

Internment camps, 133
Interpolation, fallacy of false, 122-4
Invention, history of, 33-4
Invidiam, fallacy of argument ad, 304
Ireland, history of, 50, 233
Ironsides, 245

Irrationality, and history, 187-215
Irrelevant proof, fallacy of, 45-7
Islam, 185

Israel, history of, 147
Italy, history of, 52, 74
Ivan III, 200

Jackson, Andrew, xx, 60-1, 72, 107,

110, 139, 212
Jackson, John, 273n
Jackson, Kenneth T., 240
Jacksonian democracy, 10, 60-1, 107,

113-4
Jacquerie, 245
Jaeger, Werner, 232n
James II, 83

James, William, 66, 175, 209, 213

Japan, history of, 246, 294

Jargon, 285

Jaspers, Karl, 214n

Jay, John, 105

Jay-Gardoqui treaty, 78

Jazz, 245

Jefferson, Thomas, 18, 19, 45, 96, 109-

10, 126, 133-4, 139, 140, 152, 153,

183, 246, 273, 298
Jeffersonian democracy, 100, 245; era,

116; party, 110, 125-6, 152, 202
Jeffreys, George, 288
Jensen, Merrill, 30, 31
Jeremiad, 245
Jevons, William S., 289
Jews, history of, 75, 94, 144, 147,

233, 303, 317
John Birch Society, 94
Johns Hopkins University, 26-7, 52
Johnson, Andrew, 9, 20, 182-3
Johnson, Lyndon B., 75, 77
Johnson, Samuel, 7, 225, 282, 292, 310
Johnson, Thomas H., 278
Johnson, W. S., 106
Jones, Ernest, 188n
Jordan, Terry G., 118
Jordan, Winthrop, 141, 149n
Joseph, H. W. B., 4n
Journal of Contemporary History, 218
Journalism and history, 71, 146
Judas, 245
Just, Ward S., 71n

328

INDEX

Kafka, Franz, 164

Kanl. Immanuel, 57, 66, 221

Kaplan, Abraham, xixn, xx, xxin, 37,

91, 118
Kelvin, Lord, 90, 244
Kennedy, John F., 132, 139, 168, 246
Kennedy, Robert F., 276-7
Kenny, Courtney S., 28n
Kepler, Johannes, 161
Key, V. O., 108

Keynes, John Maynard, xii, xv, 46
Khcsanh. battle of, 171
Khrushchev, Nikita, 246
Kilgallen, Dorothy, 71
Kinglake, Alexander William, 300
King Philip's War, 171
Kinship groups, 217
Kipling, Rudyard, 7. 196
Kirkland, Edward, 17n
Kirkpatrick, Jane, 6 In
Kittredge, George L., 224n
Klein, Herbert S„ 58
Kluckhohn, Clyde, 2l7n
Knollenberg, Bernhard, 176
Know-nothings, 245
Knox, Henry, 105
Koch, Adrienne, 202
Koelln, Fritz C. A., 22 In
Komarovsky, Mirra, 37n, I68n
KomrofT, Manuel, 71 n
Konigsburger, Hans, 22n
Korea, history of, 209, 250
Korean War, 250
Kremlin, 74
Krieger, Leonard, 174
Kriegspiele, 20-21
Kroeber, A. L., 2l7n
Kronenbcrger, Louis, 98
Kuba, 51

Kuhn, Thomas S., 136-7, 161-2, 2l8n
Ku Klux Klan. 240
Kulturkampf, 245

Labaree, Benjamin, 14-15
La Chapelle aux-Saintes, 110
Lamb, Harold, 72n
Lancaster, Bruce, 73n
Landon, Alf, 106-7, 271
Lane, Frederic C, 209
Lang, David Marshall, 299
Langer, William L , 188
Language groups, 217
Lanier, Henry Wysham, I66n
Lansing, John Jr., 106
Larrabee, Harold A., 169n, 282n
Laski, Harold, 269
Laslett, Peter, 59

Latin America, history of, 31; see also

specific countries and colonies
Lavender, David, 72n

Law, and history, 53, 54

Law, history of, 142

Law of averages, 118

Law of the excluded middle, 10, 12

Lawrence, T. E., 145

Leach, Douglas E., 1 7 1 n

Leach, F. R , 299

Lcbensraum, 245

Lee, Robert E., 16, 134, 173, 212-13
Lee, R. H., 105
l.efebvre, Georges, 103, 304
Leibnitz, Gottfried, 221
Lcmisch, Jesse, 208
Length, misuse of, 285, 287-8
Lenin, Nicolai, 138, 174
l.erner, Daniel, 95n, I82n
l.erncr, Gcrda, 274n
Lerner, Max. 192, 296
Leveller, 245
Lcvi-Strauss, Claude, 81 n
Levy, Leonard, 1 33-4
Lewis, Michael, 62n, 122
Lewis, Wyndham, 309
l exical definition, 278
Liberalism, and history, 26, 85, 139,
190, 207

Lincoln, Abraham, 9, 14, 72, 110, 132,

153, 168, 173. 291
Lindsay, Jack, 212
Linguistics, 21
Link, Arthur S., 54n
Literalism, misplaced, 59-61
Literary Digest, 1 06

Literature, and history, 87-90, 98, 311,

312; see also Aesthetics
Litotes, 269
Litwack, Leon, I49n
Livy, xiii, 87-8
Llorente, Juan Antonio, 300
Lloyd, Henry Demarest, 46
Locke, John, 199, 221
Loco-Focos, 245

Lodge, Henry Cabot, 31, 139, 152-3, 235,

268, 298
Logan, Stephen T., 291
Logarithms, 1 1 8

Logic, and history, i-ix, xv-xvii, 65;
abductive, xvi; adductive, xv, xvi;
canonical, 263; dichotomous, 263;
deductive, xv, xvi, 128, 267, 463-4;
of tacit inference, xvi; field-related,
ix; historical, xv; inductive,
xv-xvi, 4; mathematical, 263; syllo-
gistic, 181, 263; Anglo-American
hostility to, x; historians'
hostility to, ix-xi

Logroller, 245

Lokken. Roy N., 22-3

Lonely fact, fallacy of, 109-10

Lonn, Ella, 26n

Louis XI, 61

INDEX

329

Louis XIV, 138, 146, 250, 316
Louisiana Purchase, 18
Lowell, A. Lawrence, 90n
Lucan, 88
Lucian, 300
Luddites, 95

Luther, Martin, 174, 188, 301
Lyman, Theodore, 268
Lynd, Staughton, 83-5
Lynn, Kenneth, 96n

Mabie, Carleton, 133n

Macartney, C. A., 187n

Maculay, Catherine, 292

Macaulay, Thomas, 57, 65, 67, 271, 288.

297, 300
Macbird, 75

MacDougall, William, 213
Mach, Ernst, 164
Machiavelli, Niccold, 33
Machiavellianism, 245
Maclver, R. M., 178-9, 185, 242
Mackie, John Duncan, xn, 15 In, 177n
Mackiewicz, Stanislaw, 247
Macmillan, Harold, 159
MacMullen, Ramsay, 212, 218n
Macpherson, C. B., 199-200
Madison, James, 27, 56, 105, 106, 117
Mafia, 75
Magic, 210

Magic, and history, 218
Magna Charta, 127
Mahan, Captain, 170
Main, Jackson T., 92-3, 106
Mainstream of America series, 71
Maistre, Joseph de, 304
Maitland, Frederic, 145, 266-7
Malaya, history of, 250
Malinowski, Bronislaw, 74n
Malone, Dumas, 299
Manchester, William, 71
Manchu dynasty, 177
Man-mass, fallacy of, 208-9
Mann, Thomas, 42
Manners, history of, 230
Mannheim, Karl, 42n
Manors, 217

Many questions, fallacy of, 8-12
Mao Tse-tung, 249
March on Rome, 52
Marcuse, Herbert, 86, 195n
Marlborough, Duchess of, 174
Marlborough, Duke of, 288
Marshall, James, 72
Marshall, John, 58, 72
Marshall, T. H., 37, 38n
Martin, Luther, 77, 106
Marx, Karl, 4n, 31, 66, 83, 194n, 195n,
226

Marxism, 4, 26, 174, 194, 215, 226, 245

Maryland Historical Society, 141
Maslow, Abraham, 189, 213-14
Mason, George, 105
Mass man, fallacy of, 207-08
Materialism, historical, 1, 194;

see also Marxism
Mathematics, and history, 115
Mathematical symbols, misuse of, 289-90
Mather, Cotton, 225
Mathiez, Albert, 304
Matthews, Mitford, 278
Mattingly, Garrett, 122n, 166-7
Maung Htin Aung, 228n, 229
Maung Maung, 228n
Maurras, Charles, 304
Maxwell, James Clerk, 244
May, Ernest, 108n
Mayer, Arno J., 248-50
Mazlish, Bruce, 33, 193n, 203, 257-8
Mazzeo, Joseph A., 255n
McCallum, Francis T. and Henry D., 141n
McCarthy, Eugene, 276-7
McCarthy, Joseph, 291
McClellan, George B„ 173
McClelland, David C, 189, 213-14
McCormick, Richard, 123
McCoy, Donald R., 107n
McDonald, Forrest, 29-30, 77-8
McKeon, Richard, 87n
McKinley, William, 32, 168
McKitrick, Eric, 20-1, 27, 76n